CA2419988C - Method and apparatus for transferring video frame in telecommunication system - Google Patents
Method and apparatus for transferring video frame in telecommunication system Download PDFInfo
- Publication number
- CA2419988C CA2419988C CA002419988A CA2419988A CA2419988C CA 2419988 C CA2419988 C CA 2419988C CA 002419988 A CA002419988 A CA 002419988A CA 2419988 A CA2419988 A CA 2419988A CA 2419988 C CA2419988 C CA 2419988C
- Authority
- CA
- Canada
- Prior art keywords
- block
- coded
- reference frame
- motion
- decoded
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired - Lifetime
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- H04N19/10—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding
- H04N19/134—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the element, parameter or criterion affecting or controlling the adaptive coding
- H04N19/146—Data rate or code amount at the encoder output
- H04N19/147—Data rate or code amount at the encoder output according to rate distortion criteria
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- H04N19/10—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding
- H04N19/189—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the adaptation method, adaptation tool or adaptation type used for the adaptive coding
- H04N19/19—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the adaptation method, adaptation tool or adaptation type used for the adaptive coding using optimisation based on Lagrange multipliers
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- H04N19/50—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using predictive coding
- H04N19/503—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using predictive coding involving temporal prediction
- H04N19/51—Motion estimation or motion compensation
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- H04N19/50—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using predictive coding
- H04N19/503—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using predictive coding involving temporal prediction
- H04N19/51—Motion estimation or motion compensation
- H04N19/56—Motion estimation with initialisation of the vector search, e.g. estimating a good candidate to initiate a search
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- H04N19/50—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using predictive coding
- H04N19/503—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using predictive coding involving temporal prediction
- H04N19/51—Motion estimation or motion compensation
- H04N19/573—Motion compensation with multiple frame prediction using two or more reference frames in a given prediction direction
Landscapes
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Signal Processing (AREA)
- Compression Or Coding Systems Of Tv Signals (AREA)
- Compression, Expansion, Code Conversion, And Decoders (AREA)
- Color Television Systems (AREA)
Abstract
A method of coding video frames in a telecommunication system, comprising:
forming a video frame of consecutive stationary frames, storing the frame reconstruction data of at least one frame as a reference frame and the motion data of earlier coded neighbouring blocks, defining by means of the motion data of one or more earlier coded neighbouring blocks the motion data of the block to be coded, which neighbouring block is formed by means of the stored reference frame, defining the frame reconstruction data and motion data representing the block to be coded, which provide a pre-defined coding efficiency with a predefined picture quality.
forming a video frame of consecutive stationary frames, storing the frame reconstruction data of at least one frame as a reference frame and the motion data of earlier coded neighbouring blocks, defining by means of the motion data of one or more earlier coded neighbouring blocks the motion data of the block to be coded, which neighbouring block is formed by means of the stored reference frame, defining the frame reconstruction data and motion data representing the block to be coded, which provide a pre-defined coding efficiency with a predefined picture quality.
Description
METHOD AND APPARATUS FOR TRANSFERRING VIDEO FRAME IN
TELECOMMUNICATION SYSTEM
FIELD OF THE INVENTION
The invention relates to a method and an apparatus for coding and decoding a video frame in a telecommunication system, when the video frame is made up of a string of consecutive stationary frames, in which the frames to be transmitted are typically divided into frame blocks or segments, for instance pixel groups, specific for each frame and the data in each frame block typically comprises information indicating the luminance, colour and location of the frame block.
BACKGROUND OF THE INVENTION
Transferring video frames in telecommunication systems, such as video phones, video conference systems or Internet connections, is a demanding task due to the large amount of data needed to transfer a video frame, since the more bits are needed to transfer the data, the higher the data transmission rate must be. Various methods have been developed to solve this problem. To transfer a frame, the frame is usually divided into frame blocks whose size is selected to suit the system. Frame block information generally comprises information on the luminance, colour and location of the frame block in the frame itself. The frame block data is compressed by each block using a desired coding method. The compression is based on removing the less significant data. Compression methods are divided primarily into three classes: spectral redundancy reduction, spatial redundancy reduction and temporal redundancy reduction. Various combinations of these methods are typically used in compression.
A YUV colour model, for instance, is applied to spectral redundancy reduction. The YUV colour model utilises the fact that the human eye is more sensitive to changes in luminance than in chrominance, i.e. colour. The YUV
model has one luminance component (Y) and two chrominance components (U, V). The chrominance components are also sub-sampled. For instance, a luminance block according to the H.263 video coding standard is 16 x 16 pixels and both chrominance blocks, which cover the same area as the luminance block, are 8 x 8 pixels. The combination of one luminance block and two chrominance blocks is in this standard called a macro block.
TELECOMMUNICATION SYSTEM
FIELD OF THE INVENTION
The invention relates to a method and an apparatus for coding and decoding a video frame in a telecommunication system, when the video frame is made up of a string of consecutive stationary frames, in which the frames to be transmitted are typically divided into frame blocks or segments, for instance pixel groups, specific for each frame and the data in each frame block typically comprises information indicating the luminance, colour and location of the frame block.
BACKGROUND OF THE INVENTION
Transferring video frames in telecommunication systems, such as video phones, video conference systems or Internet connections, is a demanding task due to the large amount of data needed to transfer a video frame, since the more bits are needed to transfer the data, the higher the data transmission rate must be. Various methods have been developed to solve this problem. To transfer a frame, the frame is usually divided into frame blocks whose size is selected to suit the system. Frame block information generally comprises information on the luminance, colour and location of the frame block in the frame itself. The frame block data is compressed by each block using a desired coding method. The compression is based on removing the less significant data. Compression methods are divided primarily into three classes: spectral redundancy reduction, spatial redundancy reduction and temporal redundancy reduction. Various combinations of these methods are typically used in compression.
A YUV colour model, for instance, is applied to spectral redundancy reduction. The YUV colour model utilises the fact that the human eye is more sensitive to changes in luminance than in chrominance, i.e. colour. The YUV
model has one luminance component (Y) and two chrominance components (U, V). The chrominance components are also sub-sampled. For instance, a luminance block according to the H.263 video coding standard is 16 x 16 pixels and both chrominance blocks, which cover the same area as the luminance block, are 8 x 8 pixels. The combination of one luminance block and two chrominance blocks is in this standard called a macro block.
2 To reduce spatial redundancy, a discrete cosine transformation (DCT), for instance, is used, in which the block to be compressed is preferably 8 x 8 pixels. In a DCT, the pixel presentation of a frame block is transformed into a space-frequency presentation. In addition, only the signal frequencies which exist in a frame block have high-amplitude coefficients, whereas the coefficients of the signals that do not exist in a frame block are close to zero. DCT is also a lossless transformation and interference Is caused to the signal in quantizing only.
Temporal redundancy can be reduced by utilising the fact that consecutive frames usually resemble each other, so Instead of compressing each individual frame, motion data of the frame blocks is generated. The basic principle Is as follows: an as good as possible reference block which has been coded earlier is searched for the frame block to be coded, the motion between the reference block and the frame block to be coded is modelled and the calculated motion vector coefricients are transmitted to the receiver. The difference between the block to be coded and the reference block is indicated as a prediction error component or frame. The problem is to find a reference block which produces as good a coding efficiency as possible (a sufflciently good picture quality with as small an amount of bits as possible) and a high calculation capacity, and thus also calculation time, required by the comparison.
BRIEF DESCRIPTION OF THE INVENTION
It is thus an object of the invention to provide an improved method fnr coding video frames, an improved video coder and an improved subscriber terminal. The invention is based on using the frame reconstruction data of at least one or more reference frames or reference blocks and of one or more eariier coded neighbouring blocks in coding the frame reconstruction data. The frame reconstruction data and motion data coded with the method of the invention or with another known method is selected for use depending on how the desired coding efficiency is obtained with a sufficient picture quality.
Temporal redundancy can be reduced by utilising the fact that consecutive frames usually resemble each other, so Instead of compressing each individual frame, motion data of the frame blocks is generated. The basic principle Is as follows: an as good as possible reference block which has been coded earlier is searched for the frame block to be coded, the motion between the reference block and the frame block to be coded is modelled and the calculated motion vector coefricients are transmitted to the receiver. The difference between the block to be coded and the reference block is indicated as a prediction error component or frame. The problem is to find a reference block which produces as good a coding efficiency as possible (a sufflciently good picture quality with as small an amount of bits as possible) and a high calculation capacity, and thus also calculation time, required by the comparison.
BRIEF DESCRIPTION OF THE INVENTION
It is thus an object of the invention to provide an improved method fnr coding video frames, an improved video coder and an improved subscriber terminal. The invention is based on using the frame reconstruction data of at least one or more reference frames or reference blocks and of one or more eariier coded neighbouring blocks in coding the frame reconstruction data. The frame reconstruction data and motion data coded with the method of the invention or with another known method is selected for use depending on how the desired coding efficiency is obtained with a sufficient picture quality.
3 Accordingly, in one aspect of the present invention there is provided a method of encoding a video sequence, the method comprising:
performing a prediction for a block to be coded using one of a plurality of reference frames, the prediction comprising motion data and a reference frame number;
comparing the reference frame number of the block to be coded to the reference frame numbers of the previously coded blocks; and in response to at least one previously coded block having a different reference frame number than the reference frame number of the block to be coded, predicting the motion data of the block to be coded using only the motion data of the previously coded blocks which have the same reference frame number as the block to be coded.
According to another aspect of the present invention there is provided a terminal comprising a video codec arranged to perform a method of encoding a video sequence, the method comprising:
performing a prediction for a block to be coded using one of a plurality of reference frames, the prediction comprising motion data and a reference frame number;
comparing the reference frame number of the block to be coded to the reference frame numbers of the previously coded blocks; and in response to at least one previously coded block having a different reference frame number than the reference frame number of the block to be coded, predicting the motion data of the block to be coded using only the motion data of the previously coded blocks which have the same reference frame number as the block to be coded.
According to yet another aspect of the present invention there is provided a video encoder comprising:
a motion prediction block for performing a prediction for a block to be coded using one of a plurality of reference frames, the prediction comprising motion data and a reference frame number; and a motion coding block for comparing the reference frame number of the block to be coded to the reference frame numbers of the previously coded blocks, wherein
performing a prediction for a block to be coded using one of a plurality of reference frames, the prediction comprising motion data and a reference frame number;
comparing the reference frame number of the block to be coded to the reference frame numbers of the previously coded blocks; and in response to at least one previously coded block having a different reference frame number than the reference frame number of the block to be coded, predicting the motion data of the block to be coded using only the motion data of the previously coded blocks which have the same reference frame number as the block to be coded.
According to another aspect of the present invention there is provided a terminal comprising a video codec arranged to perform a method of encoding a video sequence, the method comprising:
performing a prediction for a block to be coded using one of a plurality of reference frames, the prediction comprising motion data and a reference frame number;
comparing the reference frame number of the block to be coded to the reference frame numbers of the previously coded blocks; and in response to at least one previously coded block having a different reference frame number than the reference frame number of the block to be coded, predicting the motion data of the block to be coded using only the motion data of the previously coded blocks which have the same reference frame number as the block to be coded.
According to yet another aspect of the present invention there is provided a video encoder comprising:
a motion prediction block for performing a prediction for a block to be coded using one of a plurality of reference frames, the prediction comprising motion data and a reference frame number; and a motion coding block for comparing the reference frame number of the block to be coded to the reference frame numbers of the previously coded blocks, wherein
4 in response to at least one previously coded block having a different reference frame number than the reference frame number of the block to be coded, the motion coding block being configured to predict the motion data of the block to be coded using only the motion data of the previously coded blocks which have the same reference frame number as the block to be coded.
According to yet another aspect of the present invention there is provided a computer-readable medium storing computer readable code for implementing a method for encoding video frames, the method comprising:
performing a prediction for a block to be coded using one of a plurality of reference frames, the prediction comprising motion data and a reference frame number;
comparing the reference frame number of the block to be coded to the reference frame numbers of the previously coded blocks; and in response to at least one previously coded block having a different reference frame number than the reference frame number of the block to be coded, predicting the motion data of the block to be coded using only the motion data of the previously coded blocks which have the same reference frame number as the block to be coded.
According to yet another aspect of the present invention there is provided a method of decoding an encoded video data comprising:
receiving motion prediction information for a block to be decoded in the encoded video data, the motion prediction information representing at least a reference frame number and motion difference data;
comparing the reference frame number of the block to be decoded to the reference frame numbers of the previously decoded blocks;
and in response to at least one previously decoded block having a different reference frame number than the reference frame number of the block to be decoded, predicting the motion data of the block to be decoded using only the motion data of the previously decoded blocks which have the same reference frame number as the block to be decoded.
According to yet another aspect of the present invention there is provided a video decoder for decoding an encoded video data, the decoder comprising:
a demultiplexing unit for receiving motion prediction information
According to yet another aspect of the present invention there is provided a computer-readable medium storing computer readable code for implementing a method for encoding video frames, the method comprising:
performing a prediction for a block to be coded using one of a plurality of reference frames, the prediction comprising motion data and a reference frame number;
comparing the reference frame number of the block to be coded to the reference frame numbers of the previously coded blocks; and in response to at least one previously coded block having a different reference frame number than the reference frame number of the block to be coded, predicting the motion data of the block to be coded using only the motion data of the previously coded blocks which have the same reference frame number as the block to be coded.
According to yet another aspect of the present invention there is provided a method of decoding an encoded video data comprising:
receiving motion prediction information for a block to be decoded in the encoded video data, the motion prediction information representing at least a reference frame number and motion difference data;
comparing the reference frame number of the block to be decoded to the reference frame numbers of the previously decoded blocks;
and in response to at least one previously decoded block having a different reference frame number than the reference frame number of the block to be decoded, predicting the motion data of the block to be decoded using only the motion data of the previously decoded blocks which have the same reference frame number as the block to be decoded.
According to yet another aspect of the present invention there is provided a video decoder for decoding an encoded video data, the decoder comprising:
a demultiplexing unit for receiving motion prediction information
5 for a block to be decoded in the encoded video data, the motion prediction information representing at least a reference frame number and motion difference data;
a motion prediction block for comparing the reference frame number of the block to be decoded to the reference frame numbers of the previously decoded blocks and for predicting the motion data of the block to be decoded using only the motion data of the previously decoded blocks which have the same reference frame number as the block to be decoded in response to at least one previously decoded block having a different reference frame number than the reference frame number of the block to be decoded.
According to yet another aspect of the present invention there is provided a terminal comprising a video codec arranged to perform a method of decoding an encoded video data, the method comprising:
receiving motion prediction information for a block to be decoded in the encoded video data, the motion prediction information representing at least a reference frame number and motion difference data;
comparing the reference frame number of the block to be decoded to the reference frame numbers of the previously decoded blocks;
and in response to at least one previously decoded block having a different reference frame number than the reference frame number of the block to be decoded, predicting the motion data of the block to be decoded using only the motion data of the previously decoded blocks which have the same reference frame number as the block to be decoded.
According to yet another aspect of the present invention there is provided a video encoder comprising:
means for performing a prediction for a block to be coded using one of a plurality of reference frames, the prediction comprising motion data and a reference frame number;
5a means for comparing the reference frame number of the block to be coded to the reference frame numbers of the previously coded blocks;
and means for predicting the motion data of the block to be coded using only the motion data of the previously coded blocks which have the same reference frame number as the block to be coded in response to at least one previously coded block having a different reference frame number than the reference frame number of the block to be coded.
According to yet another aspect of the present invention there is provided a video decoder for decoding an encoded video data, the decoder comprising:
means for receiving motion prediction information for a block to be decoded in the encoded video data, the motion prediction information representing at least a reference frame number and motion difference data;
means for comparing the reference frame number of the block to be decoded to the reference frame numbers of the previously decoded blocks; and means for predicting the motion data of the block to be decoded using only the motion data of the previously decoded blocks which have the same reference frame number as the block to be decoded in response to at least one previously coded block having a different reference frame number than the reference frame number of the block to be decoded.
According to yet another aspect of the present invention there is provided a computer-readable medium storing computer readable code for implementing a method for decoding an encoded video data, the method comprising:
receiving motion prediction information for a block to be decoded in the encoded video data, the motion prediction information representing at least a reference frame number and motion difference data;
comparing the reference frame number of the block to be decoded to the reference frame numbers of the previously decoded blocks;
and 5b in response to at least one previously decoded block having a different reference frame number than the reference frame number of the block to be decoded, predicting the motion data of the block to be decoded using only the motion data of the previously decoded blocks which have the same reference frame number as the block to be decoded.
According to yet another aspect of the present invention there is provided a method for encoding an image of a video sequence, the method comprising:
predicting a motion vector for a block to be coded based only on motion vectors of previously coded neighboring blocks which have the same reference frame used for predicting the block to be coded in response to at least one previously coded neighboring block having a different reference frame than the reference frame used for the block to be coded.
According to yet another aspect of the present invention there is provided a method for decoding an encoded image of a video sequence, the method comprising:
predicting a motion vector for a block to be decoded based only on motion vectors of previously decoded neighboring blocks which have the same reference frame used for predicting the block to be decoded in response to at least one previously decoded neighboring block having a different reference frame than the reference frame used for the block to be decoded.
According to yet another aspect of the present invention there is provided an apparatus for encoding an image of a video sequence, the apparatus comprising:
a motion coding block for predicting a motion vector for a block to be coded based only on motion vectors of previously coded neighboring blocks which have the same reference frame used for predicting the block to be coded in response to at least one previously coded neighboring block having a different reference frame than the reference frame used for the block to be coded.
5c According to yet another aspect of the present invention there is provided an apparatus for decoding an encoded image of a video sequence, the method comprising:
a motion coding block for predicting a motion vector for a block to be decoded based only on motion vectors of previously decoded neighboring blocks which have the same reference frame used for predicting the block to be decoded in response to at least one previously decoded neighboring block having a different reference frame than the reference frame used for the block to be decoded.
According to yet another aspect of the present invention there is provided a computer-readable medium embodying a computer program for encoding an image of a video sequence, said computer program comprising:
computer program code for predicting a motion vector for a block to be coded based only on motion vectors of previously coded neighboring blocks which have the same reference frame used for predicting the block to be coded in response to at least one previously coded neighboring block having a different reference frame than the reference frame used for the block to be coded.
According to still yet another aspect of the present invention there is provided a computer-readable medium embodying a computer program for decoding an encoded image of a video sequence, said computer program comprising:
computer program code for predicting a motion vector for a block to be decoded based only on motion vectors of previously decoded neighboring blocks which have the same reference frame used for predicting the block to be decoded in response to at least one previously decoded neighboring block having a different reference frame than the reference frame used for the block to be decoded.
The method and apparatus of the invention provide savings in calculation time and capacity while the picture quality remains adequate.
5d BRIEF DESCRIPTION OF THE FIGURES
In the following, the invention will be described by means of preferred embodiments, with reference to the attached drawings in which Figure 1 illustrates an example of a telecommunication system, Figures 2a to 2c illustrate one estimation process of vectors using a reference frame, Figure 3 illustrates one estimation process of a motion vector using neighbouring blocks, Figure 4 is a flow chart which shows method steps of finding a reference block and predicting a motion vector, Figure 5 illustrates an example of finding a reference block and predicting a motion vector, Figure 6 illustrates an example of a coder, Figure 7 illustrates an example of a decoder, Figure 8 illustrates an example of a subscriber terminal, Figure 9 illustrates a second example of finding a reference block and predicting a motion vector.
DETAILED DESCRIPTION OF THE INVENTION
Figure 1 illustrates in a simplified manner one digital data transmission system to which the solution of the invention can be applied. It is a part of a cellular radio system comprising a base station 104 which is in radio contact 108 and 110 with subscriber terminals 100 and 102 which can be terminals installed fixedly or in a vehicle or portable terminals.
Transceivers in the base station are connected to an antenna unit, with which the radio link to the subscriber terminals is implemented. The base station is further in contact with a base station controller 106 which transmits the connections of the terminals elsewhere in the network. The base station controller controls in a centralised manner several base stations connected to it. A control unit in the base station controller takes care of call control, data traffic control, mobility management, collection of statistics, and signalling.
a motion prediction block for comparing the reference frame number of the block to be decoded to the reference frame numbers of the previously decoded blocks and for predicting the motion data of the block to be decoded using only the motion data of the previously decoded blocks which have the same reference frame number as the block to be decoded in response to at least one previously decoded block having a different reference frame number than the reference frame number of the block to be decoded.
According to yet another aspect of the present invention there is provided a terminal comprising a video codec arranged to perform a method of decoding an encoded video data, the method comprising:
receiving motion prediction information for a block to be decoded in the encoded video data, the motion prediction information representing at least a reference frame number and motion difference data;
comparing the reference frame number of the block to be decoded to the reference frame numbers of the previously decoded blocks;
and in response to at least one previously decoded block having a different reference frame number than the reference frame number of the block to be decoded, predicting the motion data of the block to be decoded using only the motion data of the previously decoded blocks which have the same reference frame number as the block to be decoded.
According to yet another aspect of the present invention there is provided a video encoder comprising:
means for performing a prediction for a block to be coded using one of a plurality of reference frames, the prediction comprising motion data and a reference frame number;
5a means for comparing the reference frame number of the block to be coded to the reference frame numbers of the previously coded blocks;
and means for predicting the motion data of the block to be coded using only the motion data of the previously coded blocks which have the same reference frame number as the block to be coded in response to at least one previously coded block having a different reference frame number than the reference frame number of the block to be coded.
According to yet another aspect of the present invention there is provided a video decoder for decoding an encoded video data, the decoder comprising:
means for receiving motion prediction information for a block to be decoded in the encoded video data, the motion prediction information representing at least a reference frame number and motion difference data;
means for comparing the reference frame number of the block to be decoded to the reference frame numbers of the previously decoded blocks; and means for predicting the motion data of the block to be decoded using only the motion data of the previously decoded blocks which have the same reference frame number as the block to be decoded in response to at least one previously coded block having a different reference frame number than the reference frame number of the block to be decoded.
According to yet another aspect of the present invention there is provided a computer-readable medium storing computer readable code for implementing a method for decoding an encoded video data, the method comprising:
receiving motion prediction information for a block to be decoded in the encoded video data, the motion prediction information representing at least a reference frame number and motion difference data;
comparing the reference frame number of the block to be decoded to the reference frame numbers of the previously decoded blocks;
and 5b in response to at least one previously decoded block having a different reference frame number than the reference frame number of the block to be decoded, predicting the motion data of the block to be decoded using only the motion data of the previously decoded blocks which have the same reference frame number as the block to be decoded.
According to yet another aspect of the present invention there is provided a method for encoding an image of a video sequence, the method comprising:
predicting a motion vector for a block to be coded based only on motion vectors of previously coded neighboring blocks which have the same reference frame used for predicting the block to be coded in response to at least one previously coded neighboring block having a different reference frame than the reference frame used for the block to be coded.
According to yet another aspect of the present invention there is provided a method for decoding an encoded image of a video sequence, the method comprising:
predicting a motion vector for a block to be decoded based only on motion vectors of previously decoded neighboring blocks which have the same reference frame used for predicting the block to be decoded in response to at least one previously decoded neighboring block having a different reference frame than the reference frame used for the block to be decoded.
According to yet another aspect of the present invention there is provided an apparatus for encoding an image of a video sequence, the apparatus comprising:
a motion coding block for predicting a motion vector for a block to be coded based only on motion vectors of previously coded neighboring blocks which have the same reference frame used for predicting the block to be coded in response to at least one previously coded neighboring block having a different reference frame than the reference frame used for the block to be coded.
5c According to yet another aspect of the present invention there is provided an apparatus for decoding an encoded image of a video sequence, the method comprising:
a motion coding block for predicting a motion vector for a block to be decoded based only on motion vectors of previously decoded neighboring blocks which have the same reference frame used for predicting the block to be decoded in response to at least one previously decoded neighboring block having a different reference frame than the reference frame used for the block to be decoded.
According to yet another aspect of the present invention there is provided a computer-readable medium embodying a computer program for encoding an image of a video sequence, said computer program comprising:
computer program code for predicting a motion vector for a block to be coded based only on motion vectors of previously coded neighboring blocks which have the same reference frame used for predicting the block to be coded in response to at least one previously coded neighboring block having a different reference frame than the reference frame used for the block to be coded.
According to still yet another aspect of the present invention there is provided a computer-readable medium embodying a computer program for decoding an encoded image of a video sequence, said computer program comprising:
computer program code for predicting a motion vector for a block to be decoded based only on motion vectors of previously decoded neighboring blocks which have the same reference frame used for predicting the block to be decoded in response to at least one previously decoded neighboring block having a different reference frame than the reference frame used for the block to be decoded.
The method and apparatus of the invention provide savings in calculation time and capacity while the picture quality remains adequate.
5d BRIEF DESCRIPTION OF THE FIGURES
In the following, the invention will be described by means of preferred embodiments, with reference to the attached drawings in which Figure 1 illustrates an example of a telecommunication system, Figures 2a to 2c illustrate one estimation process of vectors using a reference frame, Figure 3 illustrates one estimation process of a motion vector using neighbouring blocks, Figure 4 is a flow chart which shows method steps of finding a reference block and predicting a motion vector, Figure 5 illustrates an example of finding a reference block and predicting a motion vector, Figure 6 illustrates an example of a coder, Figure 7 illustrates an example of a decoder, Figure 8 illustrates an example of a subscriber terminal, Figure 9 illustrates a second example of finding a reference block and predicting a motion vector.
DETAILED DESCRIPTION OF THE INVENTION
Figure 1 illustrates in a simplified manner one digital data transmission system to which the solution of the invention can be applied. It is a part of a cellular radio system comprising a base station 104 which is in radio contact 108 and 110 with subscriber terminals 100 and 102 which can be terminals installed fixedly or in a vehicle or portable terminals.
Transceivers in the base station are connected to an antenna unit, with which the radio link to the subscriber terminals is implemented. The base station is further in contact with a base station controller 106 which transmits the connections of the terminals elsewhere in the network. The base station controller controls in a centralised manner several base stations connected to it. A control unit in the base station controller takes care of call control, data traffic control, mobility management, collection of statistics, and signalling.
6 The cellular radio system can also have a connection to the public telephone network, in which case a transcoder belonging to the system transforms the different digital coding formats of speech used between the public telephone network and the cellular radio network to suit each other, for instance from the 64 kbit/s format of the fixed network to a cellular radio network format (e.g. 13 kbit/s) and vice versa.
It is obvious to a person skilled in the art that the invention can also be applied to fixed telecommunication systems, such as the Internet.
The aim in coding a video frame is to reduce the number of bits required to transfer the frame. Figures 2a to 2c show in a simplified manner a prior art coding method of motion which is based on predicting the motion vector of a frame block (the term block is also used in the text) using a reference frame stored earlier in memory. This type of coding is called inter-coding which means utilising the similarities between the frames in a string of frames in compression coding.
In the example of Figures 2a to 2c, the reference frame 206 has been divided into blocks or segments and coded earlier. Figures 2b to 2c do not show frame blocks. The reference frame can be before or after the frame to be coded in the string of frames. Figure 2a shows the current frame being coded 204 which is also divided into blocks, of which only one block 200 is shown for clarity's sake. The size of the blocks is N x M. The blocks are typically squares, i.e. N= M, but can also be rectangular, for instance, in their geometry. The current block 200 and the reference block 214 shown in Figure 2c need not be geometrically similar, because in reality, too, the filmed objects change, such as rotate or twist. The size and shape of the block is defined suitable for each situation taking into consideration the following matters, for instance: a small block size provides a better picture, but reduces the reliability of the defined motion vector, because the definition is made by means of a few pixels only. More motion vector data must also be generated, if smaller blocks are used, because then the frame must be divided into a greater number of blocks. In addition, fast algorithms in defining motion vectors-are more efficient when the block size is large. For instance, in some of the current video coding standards N = M = 16.
To estimate the motion, as matching a reference block as possible is found in the reference frame 206 for the block 200 being coded, this being 214 in Figure 2c. The suitability of the reference blocks depends on the
It is obvious to a person skilled in the art that the invention can also be applied to fixed telecommunication systems, such as the Internet.
The aim in coding a video frame is to reduce the number of bits required to transfer the frame. Figures 2a to 2c show in a simplified manner a prior art coding method of motion which is based on predicting the motion vector of a frame block (the term block is also used in the text) using a reference frame stored earlier in memory. This type of coding is called inter-coding which means utilising the similarities between the frames in a string of frames in compression coding.
In the example of Figures 2a to 2c, the reference frame 206 has been divided into blocks or segments and coded earlier. Figures 2b to 2c do not show frame blocks. The reference frame can be before or after the frame to be coded in the string of frames. Figure 2a shows the current frame being coded 204 which is also divided into blocks, of which only one block 200 is shown for clarity's sake. The size of the blocks is N x M. The blocks are typically squares, i.e. N= M, but can also be rectangular, for instance, in their geometry. The current block 200 and the reference block 214 shown in Figure 2c need not be geometrically similar, because in reality, too, the filmed objects change, such as rotate or twist. The size and shape of the block is defined suitable for each situation taking into consideration the following matters, for instance: a small block size provides a better picture, but reduces the reliability of the defined motion vector, because the definition is made by means of a few pixels only. More motion vector data must also be generated, if smaller blocks are used, because then the frame must be divided into a greater number of blocks. In addition, fast algorithms in defining motion vectors-are more efficient when the block size is large. For instance, in some of the current video coding standards N = M = 16.
To estimate the motion, as matching a reference block as possible is found in the reference frame 206 for the block 200 being coded, this being 214 in Figure 2c. The suitability of the reference blocks depends on the
7 selection criterion. The purpose of the selection criterion is typically to find a block, by means of which it is possible to minimise the number of bits required for transferring the block being coded in such a manner that the picture quality remains sufficiently good. Thus, the selection criteria can vary depending on the object of application.
The search is usually performed in the reference frame 206 in an area which is in a frame location corresponding to the block 200 being coded in the current frame 204 and in a certain area in the immediate proximity, depicted by the search area 210 in Figures 2b and 2c. By limiting the search area smaller than the entire reference frame, the time used in searching and the required calculation is reduced. The size of the search area is defined suitable for the object of application, for instance by determining on the basis of the frames of the frame string coded earlier, whether the motion is slow or fast; if the motion is slow, the search area can be limited smaller than if the motion is fast. The search is typically conducted in the surroundings of the predicted motion vector.
' The location of the frame block is in the example of Figures 2a to 2c indicated with coordinates (x, y) in the upper left corner of the block. The x coordinate indicates the location of the pixel or block in horizontal direction and the y coordinate the location of the pixel or block in vertical direction.
The coordinates of the block being coded are marked as 202. In Figure 2c, the best possible reference frame block 214 has been found in the search area 210. The coordinates (x, y) 208 are those of the location where the block 200 being coded would be if it had not moved and thus the same as the coordinates 202. They are shown for the purpose of illustration. The block 212 depicts the block 200 had it not moved and is also shown for the purpose of illustration. The motion of the frame block being coded can now be indicated as a change from the coordinates (x+u, y+v) 216 to the coordinates (x, y) 208.
In general, the coordinates (x, y) are defined as origo, so the motion vector 218 of the block being coded can be indicated as coordinates (u, v). The motion vector can also be indicated using definition of motion vector coefficients and motion models.
Figure 3 shows in a simplified manner prior art motion coding methods which are based on predicting the motion vector coefficients of a frame block being coded using earlier coded neighbouring blocks of the frame being coded. The neighbouring blocks of the frame block 304 being coded,
The search is usually performed in the reference frame 206 in an area which is in a frame location corresponding to the block 200 being coded in the current frame 204 and in a certain area in the immediate proximity, depicted by the search area 210 in Figures 2b and 2c. By limiting the search area smaller than the entire reference frame, the time used in searching and the required calculation is reduced. The size of the search area is defined suitable for the object of application, for instance by determining on the basis of the frames of the frame string coded earlier, whether the motion is slow or fast; if the motion is slow, the search area can be limited smaller than if the motion is fast. The search is typically conducted in the surroundings of the predicted motion vector.
' The location of the frame block is in the example of Figures 2a to 2c indicated with coordinates (x, y) in the upper left corner of the block. The x coordinate indicates the location of the pixel or block in horizontal direction and the y coordinate the location of the pixel or block in vertical direction.
The coordinates of the block being coded are marked as 202. In Figure 2c, the best possible reference frame block 214 has been found in the search area 210. The coordinates (x, y) 208 are those of the location where the block 200 being coded would be if it had not moved and thus the same as the coordinates 202. They are shown for the purpose of illustration. The block 212 depicts the block 200 had it not moved and is also shown for the purpose of illustration. The motion of the frame block being coded can now be indicated as a change from the coordinates (x+u, y+v) 216 to the coordinates (x, y) 208.
In general, the coordinates (x, y) are defined as origo, so the motion vector 218 of the block being coded can be indicated as coordinates (u, v). The motion vector can also be indicated using definition of motion vector coefficients and motion models.
Figure 3 shows in a simplified manner prior art motion coding methods which are based on predicting the motion vector coefficients of a frame block being coded using earlier coded neighbouring blocks of the frame being coded. The neighbouring blocks of the frame block 304 being coded,
8 which have already been coded earlier and which are used to predict the motion vector coefficients of the frame block being coded are here referred to as prediction candidates. In Figure 3, the prediction candidates are neighbouring blocks 300, 302, 306. The prediction candidates 300, 302, 306 are in the example of Figure 3 marked for identification with certain pixel locations 308, 310, 312. The prediction candidate 306 is on the left side of the block being coded and marked with the pixel location 308, the prediction candidate 300 is above the block being coded and marked with the pixel location 310 and the prediction candidate 302 is in the top right corner of the block being coded and marked with the pixel location 312. As seen from Figure 3, the pixel locations 308, 310, 312 reside in their blocks in close proximity to the block 304 being coded. It should be noted that the number and location of the neighbouring blocks with respect to the block being coded may differ from those shown in Figure 3.
In the case shown in Figure 3, all three prediction candidates are used in predicting the motion vector coefficients of the block 304 being coded.
The required motion vector coefficients can, for instance, be predicted as a median of the motion vector coefficients obtained from these prediction candidates. It is also possible to use only one prediction candidate in predicting the motion vector coefficients of the block 304, in which case the coefficients are obtained from the motion vector coefficients of the prediction block 306, the motion vector coefficients of the prediction block 300 or the motion vector coefficients of the prediction block 302.
Earlier coder.] neighbouring blocks that are farther than in the immediate proximity cd.."also be used, but in such a case it should be done taking into consideration that the difference in the motion of the blocks usually increases while the distance grows.
Inter- and intra-coding are typically combined block by block in coding video frames to obtain the best possible result. An intra-coded frame can be sent at suitable intervals to prevent the accumulation of possible errors caused by the transmission channel from one frame to another, finally causing the failure of the decoding of the frame.
The methods shown in Figures 2a to 2c and 3 can be used both in coding and decoding. The method used by the coder with the necessary frame data is transmitted to the decoder as a bit stream.
In the case shown in Figure 3, all three prediction candidates are used in predicting the motion vector coefficients of the block 304 being coded.
The required motion vector coefficients can, for instance, be predicted as a median of the motion vector coefficients obtained from these prediction candidates. It is also possible to use only one prediction candidate in predicting the motion vector coefficients of the block 304, in which case the coefficients are obtained from the motion vector coefficients of the prediction block 306, the motion vector coefficients of the prediction block 300 or the motion vector coefficients of the prediction block 302.
Earlier coder.] neighbouring blocks that are farther than in the immediate proximity cd.."also be used, but in such a case it should be done taking into consideration that the difference in the motion of the blocks usually increases while the distance grows.
Inter- and intra-coding are typically combined block by block in coding video frames to obtain the best possible result. An intra-coded frame can be sent at suitable intervals to prevent the accumulation of possible errors caused by the transmission channel from one frame to another, finally causing the failure of the decoding of the frame.
The methods shown in Figures 2a to 2c and 3 can be used both in coding and decoding. The method used by the coder with the necessary frame data is transmitted to the decoder as a bit stream.
9 It should be noted that as shown in Figure 3, the blocks can also be segmented. Segmentation is preferable in situations where different parts of the frame move in different ways, for instance the background remains the same while a single object, such as a person, a car or the like, moves. A
frame block can be segmented by dividing it in segments vertically, as in block 300, horizontally, of which there is no example, or in both directions, as in block 306. By segmentation, it is also possible to reduce the number of bits being transferred by using a suitable part of another already coded frame.
When defining the motion vector coefficients of the frame block being coded, various known models for motion estimation are utilised.
Different models are suitable for defining different motion, such as presenting vertical and horizontal motion or presenting rotation. Such models include the affine motion model in which the motion of a pixel, frame segment or block is indicated by a formula in which Ax and Ay form the motion vector dx(x, y) = ao + aIx + az y (1) oy(x, y) = bo + b,x + bzy wherein ao, a,, a2, bo, b, and b2 are motion vector coefficients and x and y are vector components, of which x indicates the location of a pixel or a block in horizontal direction and y the location of a pixel or a block in vertical direction, or the translational model in which the motion of a pixel, frame segment or block is indicated by a formula in which Ax and Ay form the motion vector Ax(x, y) = ao (2) Ay(x, y) = bo wherein ao and bo are motion vector coefficients and x and y are vector components. The used motion models are stored in the memory of both the coder and the decoder. If several motion models are used in coding, the coder selects the_ model according to the motion (e.g. with upward motion, a model indicating upward motion is used in coding) and information on the used model is transmitted to the decoder together with the motion vector coefficients.
Figure 4 is a flow chart showing method steps for video frame coding. In the following description, a frame block refers, for instance, to a block, segment or macro block according to the YUV model formed by pixels.
The execution of the method starts in block 400.
In block 402, a video frame is formed of consecutive stationary frames. To form the video frame, the stationary frames are sent preferably at a 5 speed at which the human visual system no longer distinguishes separate frames from each other.
In block 404, the coded reconstruction data of frame blocks (e.g.
colour and luminance of pixels and the prediction error frame obtained from the calculation of motion vector coefficients) are stored into memory. The
frame block can be segmented by dividing it in segments vertically, as in block 300, horizontally, of which there is no example, or in both directions, as in block 306. By segmentation, it is also possible to reduce the number of bits being transferred by using a suitable part of another already coded frame.
When defining the motion vector coefficients of the frame block being coded, various known models for motion estimation are utilised.
Different models are suitable for defining different motion, such as presenting vertical and horizontal motion or presenting rotation. Such models include the affine motion model in which the motion of a pixel, frame segment or block is indicated by a formula in which Ax and Ay form the motion vector dx(x, y) = ao + aIx + az y (1) oy(x, y) = bo + b,x + bzy wherein ao, a,, a2, bo, b, and b2 are motion vector coefficients and x and y are vector components, of which x indicates the location of a pixel or a block in horizontal direction and y the location of a pixel or a block in vertical direction, or the translational model in which the motion of a pixel, frame segment or block is indicated by a formula in which Ax and Ay form the motion vector Ax(x, y) = ao (2) Ay(x, y) = bo wherein ao and bo are motion vector coefficients and x and y are vector components. The used motion models are stored in the memory of both the coder and the decoder. If several motion models are used in coding, the coder selects the_ model according to the motion (e.g. with upward motion, a model indicating upward motion is used in coding) and information on the used model is transmitted to the decoder together with the motion vector coefficients.
Figure 4 is a flow chart showing method steps for video frame coding. In the following description, a frame block refers, for instance, to a block, segment or macro block according to the YUV model formed by pixels.
The execution of the method starts in block 400.
In block 402, a video frame is formed of consecutive stationary frames. To form the video frame, the stationary frames are sent preferably at a 5 speed at which the human visual system no longer distinguishes separate frames from each other.
In block 404, the coded reconstruction data of frame blocks (e.g.
colour and luminance of pixels and the prediction error frame obtained from the calculation of motion vector coefficients) are stored into memory. The
10 frame reconstruction data is coded according to the applied standard, for instance H.263 or MPEG. The definition of the motion vector coefficients and the prediction error frame is described above. One or more reference frames are formed of the frame reconstruction data stored in memory. In addition, the motion data of the earlier coded neighbouring blocks, i.e. the motion vector coefficients and a possibly used motion model, are stored into memory. It is most preferable to use neighbouring blocks in close proximity to the block being coded, but sometimes it is also possible to use earlier coded blocks farther away in the frame.
In block 406, the motion data of the block being coded is defined by means of earlier coded neighbouring blocks, preferably using the motion vector coefficients of one or more neighbouring blocks.
In block 408, the reconstruction data of the block being coded is defined using a reference frame, for instance.
The method can be intensified by calculating the motion data of the frame block being coded using only the neighbouring blocks which have the same reference frame that was used to code the frame reconstruction data of the block being coded. Similarities often exist between frames in the same frame string, so it is probable that either a block corresponding to the block being coded is found or a block that is similar enough is found in an earlier coded frame from which a suitable reference block has already been found for an earlier. coded neighbouring block. For instance, motion vector coefficients of one or more neighbouring blocks are used and the search area of the reference block of the block being coded is restricted in the reference frame to the environment defined by the motion vector coefficients of the neighbouring block.
In block 406, the motion data of the block being coded is defined by means of earlier coded neighbouring blocks, preferably using the motion vector coefficients of one or more neighbouring blocks.
In block 408, the reconstruction data of the block being coded is defined using a reference frame, for instance.
The method can be intensified by calculating the motion data of the frame block being coded using only the neighbouring blocks which have the same reference frame that was used to code the frame reconstruction data of the block being coded. Similarities often exist between frames in the same frame string, so it is probable that either a block corresponding to the block being coded is found or a block that is similar enough is found in an earlier coded frame from which a suitable reference block has already been found for an earlier. coded neighbouring block. For instance, motion vector coefficients of one or more neighbouring blocks are used and the search area of the reference block of the block being coded is restricted in the reference frame to the environment defined by the motion vector coefficients of the neighbouring block.
11 In a situation in which only one inter-coded neighbouring block has the same reference frame as the block being coded, the motion vector of this neighbouring block is selected as one motion vector candidate, another candidate can, for instance, be a median calculated from several neighbouring block motion vectors.
It should be noted that the block, which is used in coding the block, is coded in the same manner (inter) with which the block being coded will be coded, i.e., if the neighbouring block is intra-coded, it has no reference frame.
If there is no prediction candidate having a reference frame, prior art intra-coding can be used or all reference frames stored in memory can be checked.
A reference block is the frame block that the video coder, i.e. coder with which the frame reconstruction data of the frame block is coded, uses in compression coding. A frame block, which is as matching as possible with the block being coded and which was searched from one or more reference frames that were coded earlier and stored in memory, is selected as the reference block, whereby the prediction error frame remains as small as possible.
In block 410, the coding method is selected, with which the frame reconstruction data and motion data of the frame block being coded can be coded most advantageously, i.e. a pre-defined coding efficiency is reached with a pre-defined picture quality. It should be noted that as the coding method of the block being coded either the above-mentioned coding by means of the motion data of the neighbouring blocks is selected, or any prior art method, such as inter-coding, in which the block being coded is coded directly by means of the reference block found from the reference frame. In selecting the coding method, the aim may, for instance, be to make the number of bits as small as possible while the quality of the picture corresponds to that of a normal video picture or to make the number of bits match that of the used standard, even though a smaller number of bits was enough to achieve a picture quality according to the application object.
Coding efficiency can be estimated by means of a Lagrangian function, for instance, which defines the most effective segmentation and coding and which is L(S4-) = D(Sk) + a(Sk) (3) wherein
It should be noted that the block, which is used in coding the block, is coded in the same manner (inter) with which the block being coded will be coded, i.e., if the neighbouring block is intra-coded, it has no reference frame.
If there is no prediction candidate having a reference frame, prior art intra-coding can be used or all reference frames stored in memory can be checked.
A reference block is the frame block that the video coder, i.e. coder with which the frame reconstruction data of the frame block is coded, uses in compression coding. A frame block, which is as matching as possible with the block being coded and which was searched from one or more reference frames that were coded earlier and stored in memory, is selected as the reference block, whereby the prediction error frame remains as small as possible.
In block 410, the coding method is selected, with which the frame reconstruction data and motion data of the frame block being coded can be coded most advantageously, i.e. a pre-defined coding efficiency is reached with a pre-defined picture quality. It should be noted that as the coding method of the block being coded either the above-mentioned coding by means of the motion data of the neighbouring blocks is selected, or any prior art method, such as inter-coding, in which the block being coded is coded directly by means of the reference block found from the reference frame. In selecting the coding method, the aim may, for instance, be to make the number of bits as small as possible while the quality of the picture corresponds to that of a normal video picture or to make the number of bits match that of the used standard, even though a smaller number of bits was enough to achieve a picture quality according to the application object.
Coding efficiency can be estimated by means of a Lagrangian function, for instance, which defines the most effective segmentation and coding and which is L(S4-) = D(Sk) + a(Sk) (3) wherein
12 D(Sk) is an error created in despreading the frame, R(Sk) is the number of bits required to transfer the frame, and A is the Lagrangian constant.
The Lagrangian function depicts the connection between the quality of the frame being transferred and the bandwidth required on the transmission path; in general, the higher quality picture, the bigger the bandwidth. The selection of the most advantageous compression coding is always a compromise between the required number of bits and the picture quality.
Arrow 412 depicts the repeatability of the method block by block until the frame is coded. It should be noted that the method is preferably repeated until all frames in the string of frames are coded.
The execution of the method ends in block 414.
Figure 5 illustrates an example of predicting motion vectors and finding a reference block, when only the neighbouring blocks are used, which have the same reference frame that was used in coding the frame reconstruction data of the block being coded. It should be noted that that in coding the block, prior art methods are usually also used. In the end, the coding method is selected for use, which provides the desired coding efficiency with a sufficient picture quality.
The frame page 500 being coded is divided into blocks, of which only a few are shown for clarity's sake. Block 506 is the one being coded.
Frames 502, 504, 532 and 508 are reference frames stored in memory. The reference frame of blocks 510 and 514 of the earlier coded neighbouring blocks, marked as number 1, is the frame 502. The block 506 being coded and the reference blocks 510 and 514 have not moved in the frame 500 in comparison with the frame 502. The reference frame of the neighbouring block 512, marked as number 2 in the figure, is the frame 504 and the reference frame of the neighbouring block 516, marked as number 4 in the figure, is the frame 508. The blocks in the frame 500 have moved in comparison with the frames 504 and 508. The frame 532 has not been used in coding any of the neighbouring blocks of the block 506 being coded, i.e. it is not a reference frame of any neighbouring block, so the frame 532 is not used in this example to find a reference block for the block 506 being coded.
Next, the neighbouring blocks 510 and 514 are used as prediction candidates of the motion vectors of the block 506 being coded. A check is
The Lagrangian function depicts the connection between the quality of the frame being transferred and the bandwidth required on the transmission path; in general, the higher quality picture, the bigger the bandwidth. The selection of the most advantageous compression coding is always a compromise between the required number of bits and the picture quality.
Arrow 412 depicts the repeatability of the method block by block until the frame is coded. It should be noted that the method is preferably repeated until all frames in the string of frames are coded.
The execution of the method ends in block 414.
Figure 5 illustrates an example of predicting motion vectors and finding a reference block, when only the neighbouring blocks are used, which have the same reference frame that was used in coding the frame reconstruction data of the block being coded. It should be noted that that in coding the block, prior art methods are usually also used. In the end, the coding method is selected for use, which provides the desired coding efficiency with a sufficient picture quality.
The frame page 500 being coded is divided into blocks, of which only a few are shown for clarity's sake. Block 506 is the one being coded.
Frames 502, 504, 532 and 508 are reference frames stored in memory. The reference frame of blocks 510 and 514 of the earlier coded neighbouring blocks, marked as number 1, is the frame 502. The block 506 being coded and the reference blocks 510 and 514 have not moved in the frame 500 in comparison with the frame 502. The reference frame of the neighbouring block 512, marked as number 2 in the figure, is the frame 504 and the reference frame of the neighbouring block 516, marked as number 4 in the figure, is the frame 508. The blocks in the frame 500 have moved in comparison with the frames 504 and 508. The frame 532 has not been used in coding any of the neighbouring blocks of the block 506 being coded, i.e. it is not a reference frame of any neighbouring block, so the frame 532 is not used in this example to find a reference block for the block 506 being coded.
Next, the neighbouring blocks 510 and 514 are used as prediction candidates of the motion vectors of the block 506 being coded. A check is
13 made to see whether the coding type (inter) of the prediction candidates is the same as the one to be used in coding the current block. If this is so, the block 526 is used as a reference block and the motion vector coefficients are obtained by calculating an average or median of the motion vector coefficients which have been obtained by comparing the location of the blocks 510 and 518, from which the motion vector coefficients have been obtained for the neighbouring block 510, and the location of the blocks 514 and 520, from which the motion vector coefficients have been obtained for the neighbouring block 514. In addition, a prediction error frame is defined by comparing the block 506 being coded with the reference block 526. In the example of Figure 5, the block location in frames 500 and 502 has not changed, so the motion vector coefficients are not defined.
Next, the coding type of the neighbouring block 512 is checked. If it is the same as the one to be used in the current block 506, a reference block is searched for in frame 504 and the motion vector coefficients obtained from the comparison of the blocks 512 and 522 are used as the motion vector coefficients of the block 506. The prediction error frame is then defined by means of the reference block 528.
Next, the coding type of the neighbouring block 516 is checked. If it is the same as the one to be used in the current block 506, a reference block is searched for in frame 508 and the motion vector coefficients obtained from the comparison of the blocks 516 and 524 are used as the motion vector coefficients of the block 506. The prediction error frame is then defined by means of the reference block 530.
The motion vector coefficients for block 506 can also be defined by calculating a median or average of the motion vector coefficients of the blocks 510, 512 and 514.
Also, if several motion models are used, the used motion model is indicated.
The block 534 is intra-coded, so it does not have a reference frame and is thus not used in coding the block 506, if a preferable coding method is found otherwise.
The number and location of the neighbouring blocks and the number of the stored reference frames may differ from what is shown in Figure 5.
Next, the coding type of the neighbouring block 512 is checked. If it is the same as the one to be used in the current block 506, a reference block is searched for in frame 504 and the motion vector coefficients obtained from the comparison of the blocks 512 and 522 are used as the motion vector coefficients of the block 506. The prediction error frame is then defined by means of the reference block 528.
Next, the coding type of the neighbouring block 516 is checked. If it is the same as the one to be used in the current block 506, a reference block is searched for in frame 508 and the motion vector coefficients obtained from the comparison of the blocks 516 and 524 are used as the motion vector coefficients of the block 506. The prediction error frame is then defined by means of the reference block 530.
The motion vector coefficients for block 506 can also be defined by calculating a median or average of the motion vector coefficients of the blocks 510, 512 and 514.
Also, if several motion models are used, the used motion model is indicated.
The block 534 is intra-coded, so it does not have a reference frame and is thus not used in coding the block 506, if a preferable coding method is found otherwise.
The number and location of the neighbouring blocks and the number of the stored reference frames may differ from what is shown in Figure 5.
14 Figure 9 illustrates a second example of predicting motion vectors and finding a reference block. In this example, the block being coded is coded by means of only one neighbouring block, if only one neighbouring block has been inter-coded and this neighbouring block has the same reference frame as the block being coded.
The frame 900 being coded is divided into blocks, of which only a few are shown for clarity's sake. Block 912 is the one being coded. Frames 902, 904 and 906 are reference frames stored in memory. The reference frame of the earlier coded neighbouring block 914 (marked as 1) is the frame 902 and the reference frame of block 910 (marked as 2) is the frame 904. The block 908 is intra-coded, so it does not have a reference frame.
The example in Figure 9 tests whether for the block 912 being coded suitable reference blocks can be found of all the reference frames stored in memory, i.e. frames 902, 904, 906 in this example. When testing whether the frame 902 has a suitable reference block, the motion vector of the block 914 is used as the motion vector prediction candidate, since the block 914 is the only one of the three neighbouring blocks, whose reference frame is the frame 902. When testing whether the frame 904 has a suitable reference block, the motion vector of the block 910 is used as the motion vector prediction candidate, since the block 910 is the only one of the three neighbouring blocks, whose reference frame is the frame 904. The frame 906 is not a reference frame for any of the neighbouring blocks, so in the case described in the example, the motion vector is predicted using a median of the blocks 910, 914 and 908, for instance, in which case the motion vector of the inter-coded block 908 is set to zero.
It should be noted that in predicting the motion vectors of the block being coded, it is possible to use not only the motion data of the neighbouring blocks and the reference frame data, but also segmentation information of the macro block, for instance, to which the block being coded belongs. The number and location of the neighbouring blocks and the number of the stored reference blocks may differ from that shown in Figure 9.
Figure 6 shows an example of a coder with which the above-mentioned method of coding video frames can be implemented. A design principle of video coders is to minimise the prediction error frame Eõ(x, y) between the block being coded Iõ(x, y) 620 and the prediction candidate (the selected reference block) Põ(x, y) 612. The prediction error frame is defined as a difference between the block being coded 620 and the prediction candidate 612, which is obtained in an adder 632, thus Eõ(x, y) = I,:(xay)-I',, (x, y) = (4) The prediction block 612 is obtained by a method according to a preferred embodiment of the invention by using a selected reference frame and motion vectors. The coefficients of motion vectors dx(x, y),Ay(x, y) are formed in the motion vector calculation block 630 as a sum of predicted motion 10 vectors and found difference coefficients. The predicted motion vector coefficients are, according to the preferred embodiment of the invention, obtained using the motion vectors of the neighbouring blocks which have the same reference frame as the block being coded. Other methods are also used in predicting motion vectors, one of them is a prior art method in which the
The frame 900 being coded is divided into blocks, of which only a few are shown for clarity's sake. Block 912 is the one being coded. Frames 902, 904 and 906 are reference frames stored in memory. The reference frame of the earlier coded neighbouring block 914 (marked as 1) is the frame 902 and the reference frame of block 910 (marked as 2) is the frame 904. The block 908 is intra-coded, so it does not have a reference frame.
The example in Figure 9 tests whether for the block 912 being coded suitable reference blocks can be found of all the reference frames stored in memory, i.e. frames 902, 904, 906 in this example. When testing whether the frame 902 has a suitable reference block, the motion vector of the block 914 is used as the motion vector prediction candidate, since the block 914 is the only one of the three neighbouring blocks, whose reference frame is the frame 902. When testing whether the frame 904 has a suitable reference block, the motion vector of the block 910 is used as the motion vector prediction candidate, since the block 910 is the only one of the three neighbouring blocks, whose reference frame is the frame 904. The frame 906 is not a reference frame for any of the neighbouring blocks, so in the case described in the example, the motion vector is predicted using a median of the blocks 910, 914 and 908, for instance, in which case the motion vector of the inter-coded block 908 is set to zero.
It should be noted that in predicting the motion vectors of the block being coded, it is possible to use not only the motion data of the neighbouring blocks and the reference frame data, but also segmentation information of the macro block, for instance, to which the block being coded belongs. The number and location of the neighbouring blocks and the number of the stored reference blocks may differ from that shown in Figure 9.
Figure 6 shows an example of a coder with which the above-mentioned method of coding video frames can be implemented. A design principle of video coders is to minimise the prediction error frame Eõ(x, y) between the block being coded Iõ(x, y) 620 and the prediction candidate (the selected reference block) Põ(x, y) 612. The prediction error frame is defined as a difference between the block being coded 620 and the prediction candidate 612, which is obtained in an adder 632, thus Eõ(x, y) = I,:(xay)-I',, (x, y) = (4) The prediction block 612 is obtained by a method according to a preferred embodiment of the invention by using a selected reference frame and motion vectors. The coefficients of motion vectors dx(x, y),Ay(x, y) are formed in the motion vector calculation block 630 as a sum of predicted motion 10 vectors and found difference coefficients. The predicted motion vector coefficients are, according to the preferred embodiment of the invention, obtained using the motion vectors of the neighbouring blocks which have the same reference frame as the block being coded. Other methods are also used in predicting motion vectors, one of them is a prior art method in which the
15 block being coded is coded directly by means of a reference frame. The group of the motion vectors of all pixels in a frame is called a motion vector field.
Because there are a great number of pixels in a frame, a frame is in practice divided into blocks in video coding and the motion information is transmitted block by block to the receiver decoder shown in Figure 7.
At least one motion vector selected from the predicted motion vectors or motion vector fields formed of them are coded in a motion vector coding block 628. Motion models well known in the art, for instance, of which examples are presented above, are used in the coding. The coded motion vectors or motion vector fields formed of them are transmitted to a multiplexer 624.
In block 618, a frame is formed of the frame blocks segment by segment or block by block. The reference block or prediction candidate is of format Põ(x, y) = R1~[x + dx(x, y), y + Xy(x, y)] (5) wherein one or more reference frames R,k(x, y) 622 are stored in memory 610 and wherein x and y are vector components, x showing the location of a pixel, frame segment or block in horizontal direction and y in vertical direction, and dx and Xy forming the motion vector of a pixel, frame segment or block.
Because there are a great number of pixels in a frame, a frame is in practice divided into blocks in video coding and the motion information is transmitted block by block to the receiver decoder shown in Figure 7.
At least one motion vector selected from the predicted motion vectors or motion vector fields formed of them are coded in a motion vector coding block 628. Motion models well known in the art, for instance, of which examples are presented above, are used in the coding. The coded motion vectors or motion vector fields formed of them are transmitted to a multiplexer 624.
In block 618, a frame is formed of the frame blocks segment by segment or block by block. The reference block or prediction candidate is of format Põ(x, y) = R1~[x + dx(x, y), y + Xy(x, y)] (5) wherein one or more reference frames R,k(x, y) 622 are stored in memory 610 and wherein x and y are vector components, x showing the location of a pixel, frame segment or block in horizontal direction and y in vertical direction, and dx and Xy forming the motion vector of a pixel, frame segment or block.
16 In block 602, the prediction error frame Eõ(x, y) 600 is coded, which typically is presented as a finite 2-dimensional series transformation, for instance by means of a discrete cosine transformation (DCT). The coefficients obtained by the DCT transformation are quantized and coded before transmitting them to the multiplexer 624 and the receiver decoder shown in Figure 7.
The coder also comprises a decoder block 606 in which the prediction error frame E,I(x, y) is decoded. This prediction error frame En(x, y) 608 is added in the adder 634 to the prediction block Põ(x, y) 612 and a decoded block In(x, y) is obtained as a result, from which a new reference block is obtained, or by combining the blocks belonging to the same page, the reference frame 614 is obtained and stored in memory 610. Thus both the transmitter video coder and the receiver video coder have the same reference frame or block for use.
In the multiplexer 624, the information containing the motion vector information is combined with the prediction error frame and this frame information 616 is transmitted to the decoder (Figure 7).
The video coder functions described above can be implemented in many ways, for instance by software executed on a processor or by hardware, such as a logic built of separate components or ASIC (application-specific integrated circuit).
Figure 7 shows an example of a decoder, by means of which the method described above for reducing temporal redundancy in video frame transfer can be implemented. A coded prediction error frame 702 and coded motion vector coefficients or motion vector field coefficients 712 of a received video signal 700 are separated from each other in a demultiplexer 710.
In the memory 716 of the decoder, there is at least one earlier coded neighbouring block and/or a reference frame Rn (x, y) 722 divided into blocks. The same frame blocks and/or reference frame are stored in the coder and decoder memory.
In block 714, a prediction candidate Pn(x, y) 718 is formed in the-same manner as in the coder block 618 (Figure 6). A decoded prediction error frame En(x, y) 706 is formed in block 704.
In an adder 720, the decoded prediction error frame Eõ(x, y) 706 and prediction candidate Põ(x, y) 718 are added together. This produces a decoded block In(x, y) 708 of the format
The coder also comprises a decoder block 606 in which the prediction error frame E,I(x, y) is decoded. This prediction error frame En(x, y) 608 is added in the adder 634 to the prediction block Põ(x, y) 612 and a decoded block In(x, y) is obtained as a result, from which a new reference block is obtained, or by combining the blocks belonging to the same page, the reference frame 614 is obtained and stored in memory 610. Thus both the transmitter video coder and the receiver video coder have the same reference frame or block for use.
In the multiplexer 624, the information containing the motion vector information is combined with the prediction error frame and this frame information 616 is transmitted to the decoder (Figure 7).
The video coder functions described above can be implemented in many ways, for instance by software executed on a processor or by hardware, such as a logic built of separate components or ASIC (application-specific integrated circuit).
Figure 7 shows an example of a decoder, by means of which the method described above for reducing temporal redundancy in video frame transfer can be implemented. A coded prediction error frame 702 and coded motion vector coefficients or motion vector field coefficients 712 of a received video signal 700 are separated from each other in a demultiplexer 710.
In the memory 716 of the decoder, there is at least one earlier coded neighbouring block and/or a reference frame Rn (x, y) 722 divided into blocks. The same frame blocks and/or reference frame are stored in the coder and decoder memory.
In block 714, a prediction candidate Pn(x, y) 718 is formed in the-same manner as in the coder block 618 (Figure 6). A decoded prediction error frame En(x, y) 706 is formed in block 704.
In an adder 720, the decoded prediction error frame Eõ(x, y) 706 and prediction candidate Põ(x, y) 718 are added together. This produces a decoded block In(x, y) 708 of the format
17 Iõ(x,Y) = I',t(x,Y)+En(x,Y) (6) = Rn [x + ~x(x,Y), y + XY(x,Y)] + E,t(x,Y) wherein one or more earlier coded neighbouring blocks and/or reference frame Rõ~(x,y) 722 divided into blocks is stored in memory 716 and wherein x and y are vector components, of which x indicates the location of a pixel or a block in horizontal direction and y the location of a pixel or a block in vertical direction, and Xx and Xy form the motion vector of a pixel, frame segment or block.
The video coder functions described above can be implemented in many ways, for instance by software executed on a processor or by hardware, such as a logic built of separate components or ASIC (application-specific integrated circuit).
Figure 8 illustrates one terminal of a radio system, by means of which the method of a preferred embodiment of the invention can be implemented. A transmitter of a subscriber terminal 824 and a transmitter of a radio system network element perform partly the same tasks. The terminal 824 can, for instance, be a mobile phone or microcomputer including radio parts, without limiting the invention thereto. The described terminal comprises an antenna 800 which is used to transmit and receive signals through a duplex filter. The terminal can also comprise several antennas or a multi-antenna system. The terminal also comprises a transmitter 802 which amplifies and transmits to the antenna a modulated signal, a modulator 804 which modulates according to a selected modulation method the carrier with a data signal containing desired information, a receiver 806 which amplifies the signal coming from the antenna and down-converts it to a desired intermediate frequency or directly to base band, a demodulator 808 which demodulates the received signal so that the data signal can be separated from the carrier.
The subscriber terminal also has a control block 816 which controls the operation of the different parts of the terminal, such as a camera 820 and video codec 822, and takes the necessary action to process user speech or data generated by a user, such as digital signal processing (DSP), digital-to-analogue converting and filtering. In addition, in spread-spectrum systems, such as WCDMA, the signal spectrum is spread by means of a pseudo-random spreading code in a transmitter to a broad band and despread in a
The video coder functions described above can be implemented in many ways, for instance by software executed on a processor or by hardware, such as a logic built of separate components or ASIC (application-specific integrated circuit).
Figure 8 illustrates one terminal of a radio system, by means of which the method of a preferred embodiment of the invention can be implemented. A transmitter of a subscriber terminal 824 and a transmitter of a radio system network element perform partly the same tasks. The terminal 824 can, for instance, be a mobile phone or microcomputer including radio parts, without limiting the invention thereto. The described terminal comprises an antenna 800 which is used to transmit and receive signals through a duplex filter. The terminal can also comprise several antennas or a multi-antenna system. The terminal also comprises a transmitter 802 which amplifies and transmits to the antenna a modulated signal, a modulator 804 which modulates according to a selected modulation method the carrier with a data signal containing desired information, a receiver 806 which amplifies the signal coming from the antenna and down-converts it to a desired intermediate frequency or directly to base band, a demodulator 808 which demodulates the received signal so that the data signal can be separated from the carrier.
The subscriber terminal also has a control block 816 which controls the operation of the different parts of the terminal, such as a camera 820 and video codec 822, and takes the necessary action to process user speech or data generated by a user, such as digital signal processing (DSP), digital-to-analogue converting and filtering. In addition, in spread-spectrum systems, such as WCDMA, the signal spectrum is spread by means of a pseudo-random spreading code in a transmitter to a broad band and despread in a
18 receiver, thus endeavouring to increase the capacity of the channel. The control block also performs both coding and decoding, such as channel and speech coding. In addition, the control block adapts the signal being transmitted and signalling information to match the air interface standard of the used radio system. The subscriber terminal comprises one or more memories of different type which can be parts of other blocks or separate (not shown).
In the example of Figure 8, the control block also contains a program controlling the operation of the terminal. In the example of the figure, the video codec which codes and decodes the reconstruction data of a frame is separate, but it can also be a part of the control block 816.
When transmitting frames, the camera 820 records a digital picture into the memory block 818. The camera 820 can also record the picture directly into the codec 822. The codec codes a selected frame by means of one of the above methods or by a combination of several methods. The camera 820, codec 822 and memory 818 can reside in the subscriber terminal 824 or belong to a separate camera unit 826, in which case the camera unit 826 is connected to the subscriber terminal 824 with an interface 828.
A user of the subscriber terminal can when necessary view the frame on a display 814. The coded frame reconstruction data is transmitted in the same manner as speech or text data to the telecommunication system and transmitted to another phone, for instance.
The operational blocks of the terminal described above, such as the control block and the codec block 822 required for frame coding and decoding, can be implemented in many ways, for instance by software executed on a processor or by hardware, such as a logic built of separate components or ASIC (application-specific integrated circuit). The terminal functions shown in the figure can also be combined into operational blocks in many ways differing from the figure, the codec block 822 can, for instance, be divided in to a coder and decoder.
The user interface of the terminal comprises a loudspeaker or an ear piece 810, a microphone 812, a display 814 and possibly a keyboard, which all are connected to the control block 816.
Even though the invention has been explained in the above with reference to examples in accordance with the accompanying drawings, it is obvious that the invention is not restricted to them but can be modified in many ways within the scope of the inventive idea disclosed in the attached claims.
In the example of Figure 8, the control block also contains a program controlling the operation of the terminal. In the example of the figure, the video codec which codes and decodes the reconstruction data of a frame is separate, but it can also be a part of the control block 816.
When transmitting frames, the camera 820 records a digital picture into the memory block 818. The camera 820 can also record the picture directly into the codec 822. The codec codes a selected frame by means of one of the above methods or by a combination of several methods. The camera 820, codec 822 and memory 818 can reside in the subscriber terminal 824 or belong to a separate camera unit 826, in which case the camera unit 826 is connected to the subscriber terminal 824 with an interface 828.
A user of the subscriber terminal can when necessary view the frame on a display 814. The coded frame reconstruction data is transmitted in the same manner as speech or text data to the telecommunication system and transmitted to another phone, for instance.
The operational blocks of the terminal described above, such as the control block and the codec block 822 required for frame coding and decoding, can be implemented in many ways, for instance by software executed on a processor or by hardware, such as a logic built of separate components or ASIC (application-specific integrated circuit). The terminal functions shown in the figure can also be combined into operational blocks in many ways differing from the figure, the codec block 822 can, for instance, be divided in to a coder and decoder.
The user interface of the terminal comprises a loudspeaker or an ear piece 810, a microphone 812, a display 814 and possibly a keyboard, which all are connected to the control block 816.
Even though the invention has been explained in the above with reference to examples in accordance with the accompanying drawings, it is obvious that the invention is not restricted to them but can be modified in many ways within the scope of the inventive idea disclosed in the attached claims.
19 The invention can be applied together with several video coding standards, such as H.263 or MPEG.
Claims (57)
1. A method of encoding a video sequence, the method comprising:
performing a prediction for a block to be coded using one of a plurality of reference frames, the prediction comprising motion data and a reference frame number;
comparing the reference frame number of the block to be coded to the reference frame numbers of the previously coded blocks; and in response to at least one previously coded block having a different reference frame number than the reference frame number of the block to be coded, predicting the motion data of the block to be coded using only the motion data of the previously coded blocks which have the same reference frame number as the block to be coded.
performing a prediction for a block to be coded using one of a plurality of reference frames, the prediction comprising motion data and a reference frame number;
comparing the reference frame number of the block to be coded to the reference frame numbers of the previously coded blocks; and in response to at least one previously coded block having a different reference frame number than the reference frame number of the block to be coded, predicting the motion data of the block to be coded using only the motion data of the previously coded blocks which have the same reference frame number as the block to be coded.
2. A method according to claim 1, wherein if only one previously coded block has the same reference frame number as the block to be coded, the prediction of the motion data of the block to be coded is set equal to the motion data of the previously coded block which has the same reference frame number as the block to be coded.
3. A method according to claim 1, wherein the prediction of the motion data of the block to be coded is the median of the motion data of the previously coded blocks which have the same reference frame number as the block to be coded.
4. A method according to claim 1, wherein the prediction of the motion data of the block to be coded is the average of the motion data of the previously coded blocks which have the same reference frame number as the block to be coded.
5. A method according to claim 1, further comprising:
selecting a motion model from a plurality of motion models; and signaling the selected motion model.
selecting a motion model from a plurality of motion models; and signaling the selected motion model.
6. A method according to claim 2, further comprising:
selecting a motion model from a plurality of motion models; and signaling the selected motion model.
selecting a motion model from a plurality of motion models; and signaling the selected motion model.
7. A terminal comprising a video codec arranged to perform a method of encoding a video sequence, the method comprising:
performing a prediction for a block to be coded using one of a plurality of reference frames, the prediction comprising motion data and a reference frame number;
comparing the reference frame number of the block to be coded to the reference frame numbers of the previously coded blocks; and in response to at least one previously coded block having a different reference frame number than the reference frame number of the block to be coded, predicting the motion data of the block to be coded using only the motion data of the previously coded blocks which have the same reference frame number as the block to be coded.
performing a prediction for a block to be coded using one of a plurality of reference frames, the prediction comprising motion data and a reference frame number;
comparing the reference frame number of the block to be coded to the reference frame numbers of the previously coded blocks; and in response to at least one previously coded block having a different reference frame number than the reference frame number of the block to be coded, predicting the motion data of the block to be coded using only the motion data of the previously coded blocks which have the same reference frame number as the block to be coded.
8. A terminal according to claim 7, wherein if only one previously coded block has the same reference frame number as the block to be coded, the prediction of the motion data of the block to be coded is set equal to the motion data of the previously coded block which has the same reference frame number as the block to be coded.
9. A video encoder comprising:
a motion prediction block for performing a prediction for a block to be coded using one of a plurality of reference frames, the prediction comprising motion data and a reference frame number; and a motion coding block for comparing the reference frame number of the block to be coded to the reference frame numbers of the previously coded blocks, wherein in response to at least one previously coded block having a different reference frame number than the reference frame number of the block to be coded, the motion coding block being configured to predict the motion data of the block to be coded using only the motion data of the previously coded blocks which have the same reference frame number as the block to be coded.
a motion prediction block for performing a prediction for a block to be coded using one of a plurality of reference frames, the prediction comprising motion data and a reference frame number; and a motion coding block for comparing the reference frame number of the block to be coded to the reference frame numbers of the previously coded blocks, wherein in response to at least one previously coded block having a different reference frame number than the reference frame number of the block to be coded, the motion coding block being configured to predict the motion data of the block to be coded using only the motion data of the previously coded blocks which have the same reference frame number as the block to be coded.
10. A video encoder according to claim 9, wherein if only one previously coded block has the same reference frame number as the block to be coded, the prediction of the motion data of the block to be coded is set equal to the motion data of the previously coded block which has the same reference frame number as the block to be coded.
11. A video encoder according to claim 9, wherein the prediction of the motion data of the block to be coded is the median of the motion data of the previously coded blocks which have the same reference frame number as the block to be coded.
12. A video encoder according to claim 9, wherein the prediction of the motion data of the block to be coded is the average of the motion data of the previously coded blocks which have the same reference frame number as the block to be coded.
13. A video encoder according to claim 9, wherein the motion coding block is configured to select a motion model from a plurality of motion models and the encoder further comprises:
a multiplexer for signaling the selected motion model.
a multiplexer for signaling the selected motion model.
14. A video encoder according to claim 10, wherein the motion coding block is configured to select a motion model from a plurality of motion models and the encoder further comprises:
a multiplexer for signaling the selected motion model.
a multiplexer for signaling the selected motion model.
15. A computer-readable medium storing computer readable code for implementing a method for encoding video frames, the method comprising:
performing a prediction for a block to be coded using one of a plurality of reference frames, the prediction comprising motion data and a reference frame number;
comparing the reference frame number of the block to be coded to the reference frame numbers of the previously coded blocks; and in response to at least one previously coded block having a different reference frame number than the reference frame number of the block to be coded, predicting the motion data of the block to be coded using only the motion data of the previously coded blocks which have the same reference frame number as the block to be coded.
performing a prediction for a block to be coded using one of a plurality of reference frames, the prediction comprising motion data and a reference frame number;
comparing the reference frame number of the block to be coded to the reference frame numbers of the previously coded blocks; and in response to at least one previously coded block having a different reference frame number than the reference frame number of the block to be coded, predicting the motion data of the block to be coded using only the motion data of the previously coded blocks which have the same reference frame number as the block to be coded.
16. A computer-readable medium according to claim 15, wherein if only one previously coded block has the same reference frame number as the block to be coded, the prediction of the motion data of the block to be coded is set equal to the motion data of the previously coded block which has the same reference frame number as the block to be coded.
17. A computer-readable medium according to claim 15, wherein the prediction of the motion data of the block to be coded is the median of the motion data of the previously coded blocks which have the same reference frame number as the block to be coded.
18. A computer-readable medium according to claim 15, wherein the prediction of the motion data of the block to be coded is the average of the motion data of the previously coded blocks which have the same reference frame number as the block to be coded.
19. A computer-readable medium according to claim 15, wherein the method further comprises:
selecting a motion model from a plurality of motion models; and signaling the selected motion model.
selecting a motion model from a plurality of motion models; and signaling the selected motion model.
20. A computer-readable medium according to claim 16, wherein the method further comprises:
selecting a motion model from a plurality of motion models; and signaling the selected motion model.
selecting a motion model from a plurality of motion models; and signaling the selected motion model.
21. A method of decoding an encoded video data comprising:
receiving motion prediction information for a block to be decoded in the encoded video data, the motion prediction information representing at least a reference frame number and motion difference data;
comparing the reference frame number of the block to be decoded to the reference frame numbers of the previously decoded blocks; and in response to at least one previously decoded block having a different reference frame number than the reference frame number of the block to be decoded, predicting the motion data of the block to be decoded using only the motion data of the previously decoded blocks which have the same reference frame number as the block to be decoded.
receiving motion prediction information for a block to be decoded in the encoded video data, the motion prediction information representing at least a reference frame number and motion difference data;
comparing the reference frame number of the block to be decoded to the reference frame numbers of the previously decoded blocks; and in response to at least one previously decoded block having a different reference frame number than the reference frame number of the block to be decoded, predicting the motion data of the block to be decoded using only the motion data of the previously decoded blocks which have the same reference frame number as the block to be decoded.
22. A method according to claim 21, wherein if only one previously coded block has the same reference frame number as the block to be decoded, the prediction of the motion data of the block to be decoded is set equal to the motion data of the previously coded block which has the same reference frame number as the block to be coded.
23. A method according to claim 21, wherein the prediction of the motion data of the block to be coded is the median of the motion data of the previously coded blocks which have the same reference frame number as the block to be coded.
24. A method according to claim 21, wherein the prediction of the motion data of the block to be coded is the average of the motion data of the previously coded blocks which have the same reference frame number as the block to be coded.
25. A method according to claim 21, further comprising:
determining a motion model from a plurality of motion models; and forming a prediction for the block to be decoded using the motion model and the motion data.
determining a motion model from a plurality of motion models; and forming a prediction for the block to be decoded using the motion model and the motion data.
26. A method according to claim 25, wherein the motion model is determined by a signaling in the encoded video data.
27. A method according to claim 22, further comprising:
determining a motion model from a plurality of motion models; and forming a prediction for the block to be decoded using the motion model and the motion data.
determining a motion model from a plurality of motion models; and forming a prediction for the block to be decoded using the motion model and the motion data.
28. A method according to claim 27, wherein the motion model is determined by a signaling in the encoded video data.
29. A video decoder for decoding an encoded video data, the decoder comprising:
a demultiplexing unit for receiving motion prediction information for a block to be decoded in the encoded video data, the motion prediction information representing at least a reference frame number and motion difference data;
a motion prediction block for comparing the reference frame number of the block to be decoded to the reference frame numbers of the previously decoded blocks and for predicting the motion data of the block to be decoded using only the motion data of the previously decoded blocks which have the same reference frame number as the block to be decoded in response to at least one previously decoded block having a different reference frame number than the reference frame number of the block to be decoded.
a demultiplexing unit for receiving motion prediction information for a block to be decoded in the encoded video data, the motion prediction information representing at least a reference frame number and motion difference data;
a motion prediction block for comparing the reference frame number of the block to be decoded to the reference frame numbers of the previously decoded blocks and for predicting the motion data of the block to be decoded using only the motion data of the previously decoded blocks which have the same reference frame number as the block to be decoded in response to at least one previously decoded block having a different reference frame number than the reference frame number of the block to be decoded.
30. A video decoder according to claim 29, wherein if only one previously coded block has the same reference frame number as the block to be decoded, the prediction of the motion data of the block to be decoded is set equal to the motion data of the previously coded block which has the same reference frame number as the block to be coded.
31. A video decoder according to claim 29, wherein the prediction of the motion data of the block to be coded is the median of the motion data of the previously coded blocks which have the same reference frame number as the block to be coded.
32. A video decoder according to claim 29, wherein the prediction of the motion data of the block to be coded is the average of the motion data of the previously coded blocks which have the same reference frame number as the block to be coded.
33. A video decoder according to claim 29, wherein the motion prediction block is configured to determine a motion model from a plurality of motion models and to form a prediction for the block to be decoded using the motion model and the motion data.
34. A video decoder according to claim 33, wherein the motion model is determined by a signaling in the encoded video data at the demultiplexing unit.
35. A video decoder according to claim 30, wherein the motion prediction block is configured to determine a motion model from a plurality of motion models and to form a prediction for the block to be decoded using the motion model and the motion data.
36. A video decoder according to claim 35, wherein the motion model is determined by a signaling in the encoded video data at the demultiplexing unit.
37. A terminal comprising a video codec arranged to perform a method of decoding an encoded video data, the method comprising:
receiving motion prediction information for a block to be decoded in the encoded video data, the motion prediction information representing at least a reference frame number and motion difference data;
comparing the reference frame number of the block to be decoded to the reference frame numbers of the previously decoded blocks; and in response to at least one previously decoded block having a different reference frame number than the reference frame number of the block to be decoded, predicting the motion data of the block to be decoded using only the motion data of the previously decoded blocks which have the same reference frame number as the block to be decoded.
receiving motion prediction information for a block to be decoded in the encoded video data, the motion prediction information representing at least a reference frame number and motion difference data;
comparing the reference frame number of the block to be decoded to the reference frame numbers of the previously decoded blocks; and in response to at least one previously decoded block having a different reference frame number than the reference frame number of the block to be decoded, predicting the motion data of the block to be decoded using only the motion data of the previously decoded blocks which have the same reference frame number as the block to be decoded.
38. A terminal according to claim 37, wherein if only one previously coded block has the same reference frame number as the block to be coded, the prediction of the motion data of the block to be coded is set equal to the motion data of the previously coded block which has the same reference frame number as the block to be coded.
39. A video encoder comprising:
means for performing a prediction for a block to be coded using one of a plurality of reference frames, the prediction comprising motion data and a reference frame number;
means for comparing the reference frame number of the block to be coded to the reference frame numbers of the previously coded blocks; and means for predicting the motion data of the block to be coded using only the motion data of the previously coded blocks which have the same reference frame number as the block to be coded in response to at least one previously coded block having a different reference frame number than the reference frame number of the block to be coded.
means for performing a prediction for a block to be coded using one of a plurality of reference frames, the prediction comprising motion data and a reference frame number;
means for comparing the reference frame number of the block to be coded to the reference frame numbers of the previously coded blocks; and means for predicting the motion data of the block to be coded using only the motion data of the previously coded blocks which have the same reference frame number as the block to be coded in response to at least one previously coded block having a different reference frame number than the reference frame number of the block to be coded.
40. A video encoder according to claim 39, wherein if only one previously coded block has the same reference frame number as the block to be coded, the prediction of the motion data of the block to be coded is set equal to the motion data of the previously coded block which has the same reference frame number as the block to be coded.
41. A video decoder for decoding an encoded video data, the decoder comprising:
means for receiving motion prediction information for a block to be decoded in the encoded video data, the motion prediction information representing at least a reference frame number and motion difference data;
means for comparing the reference frame number of the block to be decoded to the reference frame numbers of the previously decoded blocks;
and means for predicting the motion data of the block to be decoded using only the motion data of the previously decoded blocks which have the same reference frame number as the block to be decoded in response to at least one previously coded block having a different reference frame number than the reference frame number of the block to be decoded.
means for receiving motion prediction information for a block to be decoded in the encoded video data, the motion prediction information representing at least a reference frame number and motion difference data;
means for comparing the reference frame number of the block to be decoded to the reference frame numbers of the previously decoded blocks;
and means for predicting the motion data of the block to be decoded using only the motion data of the previously decoded blocks which have the same reference frame number as the block to be decoded in response to at least one previously coded block having a different reference frame number than the reference frame number of the block to be decoded.
42. A video decoder according to claim 41, wherein if only one previously coded block has the same reference frame number as the block to be decoded, the prediction of the motion data of the block to be decoded is set equal to the motion data of the previously coded block which has the same reference frame number as the block to be coded.
43. A video decoder according to claim 41, the decoder further comprises:
means for determining a motion model from a plurality of motion models by a signaling in the encoded video data; and means forming a prediction for the block to be decoded using the motion model and the motion data.
means for determining a motion model from a plurality of motion models by a signaling in the encoded video data; and means forming a prediction for the block to be decoded using the motion model and the motion data.
44. A video decoder according to claim 42, the decoder further comprises:
means for determining a motion model from a plurality of motion models by a signaling in the encoded video data; and means forming a prediction for the block to be decoded using the motion model and the motion data.
means for determining a motion model from a plurality of motion models by a signaling in the encoded video data; and means forming a prediction for the block to be decoded using the motion model and the motion data.
45. A computer-readable medium storing computer readable code for implementing a method for decoding an encoded video data, the method comprising:
receiving motion prediction information for a block to be decoded in the encoded video data, the motion prediction information representing at least a reference frame number and motion difference data;
comparing the reference frame number of the block to be decoded to the reference frame numbers of the previously decoded blocks; and in response to at least one previously decoded block having a different reference frame number than the reference frame number of the block to be decoded, predicting the motion data of the block to be decoded using only the motion data of the previously decoded blocks which have the same reference frame number as the block to be decoded.
receiving motion prediction information for a block to be decoded in the encoded video data, the motion prediction information representing at least a reference frame number and motion difference data;
comparing the reference frame number of the block to be decoded to the reference frame numbers of the previously decoded blocks; and in response to at least one previously decoded block having a different reference frame number than the reference frame number of the block to be decoded, predicting the motion data of the block to be decoded using only the motion data of the previously decoded blocks which have the same reference frame number as the block to be decoded.
46. A computer-readable medium according to claim 45, wherein if only one previously coded block has the same reference frame number as the block to be coded, the prediction of the motion data of the block to be coded is set equal to the motion data of the previously coded block which has the same reference frame number as the block to be coded.
47. A computer-readable medium according to claim 45, wherein the method further comprises:
determining a motion model from a plurality of motion models by a signaling in the encoded video data; and forming a prediction for the block to be decoded using the motion model and the motion data.
determining a motion model from a plurality of motion models by a signaling in the encoded video data; and forming a prediction for the block to be decoded using the motion model and the motion data.
48. A method for encoding an image of a video sequence, the method comprising:
predicting a motion vector for a block to be coded based only on motion vectors of previously coded neighboring blocks which have the same reference frame used for predicting the block to be coded in response to at least one previously coded neighboring block having a different reference frame than the reference frame used for the block to be coded.
predicting a motion vector for a block to be coded based only on motion vectors of previously coded neighboring blocks which have the same reference frame used for predicting the block to be coded in response to at least one previously coded neighboring block having a different reference frame than the reference frame used for the block to be coded.
49. A method according to claim 48, wherein when only one previously coded neighboring block has the same reference frame as the block to be coded, the prediction of the motion vector of the block to be coded is set equal to the motion vector of the previously coded neighboring block which has the same reference frame as the block to be coded.
50. A method for decoding an encoded image of a video sequence, the method comprising:
predicting a motion vector for a block to be decoded based only on motion vectors of previously decoded neighboring blocks which have the same reference frame used for predicting the block to be decoded in response to at least one previously decoded neighboring block having a different reference frame than the reference frame used for the block to be decoded.
predicting a motion vector for a block to be decoded based only on motion vectors of previously decoded neighboring blocks which have the same reference frame used for predicting the block to be decoded in response to at least one previously decoded neighboring block having a different reference frame than the reference frame used for the block to be decoded.
51. A method according to claim 50, wherein when only one previously decoded neighboring block has the same reference frame as the block to be decoded, the prediction of the motion vector of the block to be decoded is set equal to the motion vector of the previously decoded neighboring block which has the same reference frame as the block to be decoded.
52. An apparatus for encoding an image of a video sequence, the apparatus comprising:
a motion coding block for predicting a motion vector for a block to be coded based only on motion vectors of previously coded neighboring blocks which have the same reference frame used for predicting the block to be coded in response to at least one previously coded neighboring block having a different reference frame than the reference frame used for the block to be coded.
a motion coding block for predicting a motion vector for a block to be coded based only on motion vectors of previously coded neighboring blocks which have the same reference frame used for predicting the block to be coded in response to at least one previously coded neighboring block having a different reference frame than the reference frame used for the block to be coded.
53. An apparatus according to claim 52, wherein when only one previously coded neighboring block has the same reference frame as the block to be coded, the prediction of the motion vector of the block to be coded is set equal to the motion vector of the previously coded neighboring block which has the same reference frame as the block to be coded.
54. An apparatus for decoding an encoded image of a video sequence, the method comprising:
a motion coding block for predicting a motion vector for a block to be decoded based only on motion vectors of previously decoded neighboring blocks which have the same reference frame used for predicting the block to be decoded in response to at least one previously decoded neighboring block having a different reference frame than the reference frame used for the block to be decoded.
a motion coding block for predicting a motion vector for a block to be decoded based only on motion vectors of previously decoded neighboring blocks which have the same reference frame used for predicting the block to be decoded in response to at least one previously decoded neighboring block having a different reference frame than the reference frame used for the block to be decoded.
55. An apparatus according to claim 54, wherein when only one previously decoded neighboring block has the same reference frame as the block to be decoded, the prediction of the motion vector of the block to be decoded is set equal to the motion vector of the previously decoded neighboring block which has the same reference frame as the block to be decoded.
56. A computer-readable medium embodying a computer program for encoding an image of a video sequence, said computer program comprising:
computer program code for predicting a motion vector for a block to be coded based only on motion vectors of previously coded neighboring blocks which have the same reference frame used for predicting the block to be coded in response to at least one previously coded neighboring block having a different reference frame than the reference frame used for the block to be coded.
computer program code for predicting a motion vector for a block to be coded based only on motion vectors of previously coded neighboring blocks which have the same reference frame used for predicting the block to be coded in response to at least one previously coded neighboring block having a different reference frame than the reference frame used for the block to be coded.
57. A computer-readable medium embodying a computer program for decoding an encoded image of a video sequence, said computer program comprising:
computer program code for predicting a motion vector for a block to be decoded based only on motion vectors of previously decoded neighboring blocks which have the same reference frame used for predicting the block to be decoded in response to at least one previously decoded neighboring block having a different reference frame than the reference frame used for the block to be decoded.
computer program code for predicting a motion vector for a block to be decoded based only on motion vectors of previously decoded neighboring blocks which have the same reference frame used for predicting the block to be decoded in response to at least one previously decoded neighboring block having a different reference frame than the reference frame used for the block to be decoded.
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| PCT/FI2000/000686 WO2002015585A1 (en) | 2000-08-11 | 2000-08-11 | Method and apparatus for transferring video frame in telecommunication system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CA2419988A1 CA2419988A1 (en) | 2002-02-21 |
| CA2419988C true CA2419988C (en) | 2009-11-17 |
Family
ID=8555876
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CA002419988A Expired - Lifetime CA2419988C (en) | 2000-08-11 | 2000-08-11 | Method and apparatus for transferring video frame in telecommunication system |
Country Status (12)
| Country | Link |
|---|---|
| EP (1) | EP1325635A1 (en) |
| JP (1) | JP2004534411A (en) |
| KR (1) | KR100828378B1 (en) |
| CN (2) | CN100591131C (en) |
| AU (1) | AU2000264472A1 (en) |
| BR (1) | BRPI0017304B1 (en) |
| CA (1) | CA2419988C (en) |
| EE (1) | EE05596B1 (en) |
| HK (1) | HK1100115A1 (en) |
| HU (1) | HU228614B1 (en) |
| MX (1) | MXPA03001270A (en) |
| WO (1) | WO2002015585A1 (en) |
Families Citing this family (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6735249B1 (en) | 1999-08-11 | 2004-05-11 | Nokia Corporation | Apparatus, and associated method, for forming a compressed motion vector field utilizing predictive motion coding |
| US8005145B2 (en) | 2000-08-11 | 2011-08-23 | Nokia Corporation | Method and apparatus for transferring video frame in telecommunication system |
| US8073048B2 (en) | 2003-08-26 | 2011-12-06 | Thomson Licensing | Method and apparatus for minimizing number of reference pictures used for inter-coding |
| JP2007312397A (en) * | 2007-05-25 | 2007-11-29 | Nokia Corp | Method and apparatus for video frame transfer in communication system |
| JP5716438B2 (en) * | 2011-02-08 | 2015-05-13 | 株式会社Ｊｖｃケンウッド | Image decoding apparatus, image decoding method, and image decoding program |
| JP5716437B2 (en) * | 2011-02-08 | 2015-05-13 | 株式会社Ｊｖｃケンウッド | Image coding apparatus, image coding method, and image coding program |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| DE19509418A1 (en) | 1995-03-16 | 1996-09-19 | Thomson Brandt Gmbh | Method and circuit arrangement for subsampling in motion estimation |
| US5682209A (en) * | 1995-11-13 | 1997-10-28 | Tektronix, Inc. | Motion estimation using limited-time early exit with prequalification matrices and a predicted search center |
| US6128047A (en) * | 1998-05-20 | 2000-10-03 | Sony Corporation | Motion estimation process and system using sparse search block-matching and integral projection |
| US6130912A (en) * | 1998-06-09 | 2000-10-10 | Sony Electronics, Inc. | Hierarchical motion estimation process and system using block-matching and integral projection |
-
2000
- 2000-08-11 CA CA002419988A patent/CA2419988C/en not_active Expired - Lifetime
- 2000-08-11 WO PCT/FI2000/000686 patent/WO2002015585A1/en active Application Filing
- 2000-08-11 AU AU2000264472A patent/AU2000264472A1/en not_active Abandoned
- 2000-08-11 JP JP2002520565A patent/JP2004534411A/en not_active Withdrawn
- 2000-08-11 EP EP00951576A patent/EP1325635A1/en not_active Withdrawn
- 2000-08-11 CN CN200610153497A patent/CN100591131C/en not_active Expired - Lifetime
- 2000-08-11 KR KR1020037001984A patent/KR100828378B1/en active IP Right Grant
- 2000-08-11 HU HU0302895A patent/HU228614B1/en active IP Right Revival
- 2000-08-11 CN CNB008199256A patent/CN1284372C/en not_active Expired - Lifetime
- 2000-08-11 BR BRPI0017304A patent/BRPI0017304B1/en active IP Right Grant
- 2000-08-11 MX MXPA03001270A patent/MXPA03001270A/en active IP Right Grant
- 2000-08-11 EE EEP200300053A patent/EE05596B1/en unknown
-
2007
- 2007-07-24 HK HK07108052.1A patent/HK1100115A1/en not_active IP Right Cessation
Also Published As
| Publication number | Publication date |
|---|---|
| CA2419988A1 (en) | 2002-02-21 |
| BR0017304A (en) | 2003-07-08 |
| EP1325635A1 (en) | 2003-07-09 |
| KR100828378B1 (en) | 2008-05-08 |
| JP2004534411A (en) | 2004-11-11 |
| WO2002015585A1 (en) | 2002-02-21 |
| EE05596B1 (en) | 2012-10-15 |
| HUP0302895A2 (en) | 2003-12-29 |
| AU2000264472A1 (en) | 2002-02-25 |
| CN1454431A (en) | 2003-11-05 |
| BRPI0017304B1 (en) | 2016-11-29 |
| HK1100115A1 (en) | 2010-06-04 |
| MXPA03001270A (en) | 2003-06-09 |
| EE200300053A (en) | 2004-12-15 |
| HU228614B1 (en) | 2013-04-29 |
| CN1933603A (en) | 2007-03-21 |
| CN1284372C (en) | 2006-11-08 |
| KR20030027021A (en) | 2003-04-03 |
| HUP0302895A3 (en) | 2004-04-28 |
| CN100591131C (en) | 2010-02-17 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US6782053B1 (en) | Method and apparatus for transferring video frame in telecommunication system | |
| US8005145B2 (en) | Method and apparatus for transferring video frame in telecommunication system | |
| JP2007135219A (en) | Method and apparatus for video frame transfer in communication system | |
| JP2007135219A6 (en) | Video frame transfer method and apparatus in communication system | |
| US8526748B2 (en) | Picture coding method and picture decoding method | |
| MXPA04012133A (en) | Spatial prediction based intra coding. | |
| KR101409826B1 (en) | Method for motion prediction using adaptive search range | |
| CN115988206B (en) | Image processing method, processing apparatus, and storage medium | |
| US20240121421A1 (en) | Motion vector obtaining method and apparatus | |
| CA2419988C (en) | Method and apparatus for transferring video frame in telecommunication system | |
| EP1227684A2 (en) | Encoding of video signals | |
| WO2002067590A1 (en) | Video encoding of still images | |
| JP2007312397A (en) | Method and apparatus for video frame transfer in communication system | |
| ZA200301088B (en) | Method and apparatus for transferring video frame in telecommunication system. | |
| WO2020181540A1 (en) | Video processing method and device, encoding apparatus, and decoding apparatus |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| EEER | Examination request | ||
| MKEX | Expiry |
Effective date: 20200811 |