KR20190126481A - Apparatus for encoding video using transmission control matrix and method thereof - Google Patents

Abstract

The present invention discloses a video encoding apparatus using a transmission control matrix and an operating method thereof. The video encoding apparatus transmits a source packet to M destinations using a transmission control matrix, classifies the packet into a state when the switching request information of an encoder is received from a destination where packet loss occurs and a state when the switching of the encoder is completed, and changes the component values of the transmission control matrix. So, the break of a video or interruption of a video flow is prevented during the conversion process of the encoder. One context switch and one copy are performed by a kernel space in a user space when the same source packet is sent to multiple destinations. So, load and power consumption are reduced. The video encoding apparatus includes an encoding part, first and second generation parts, and a communication part.

Description

Image encoding apparatus using transmission control matrix and its operation method {APPARATUS FOR ENCODING VIDEO USING TRANSMISSION CONTROL MATRIX AND METHOD THEREOF}

The present invention relates to an image encoding apparatus and an operation method thereof, and more particularly, to a technique for transmitting a source packet to M destinations using a transmission control matrix.

The patent document relates to a UWB-based communication system including one transmitter and a plurality of receivers, and generates a packet in one transmitter and transmits the generated packet to the plurality of receivers.

Conventionally, as described in the patent literature, a communication system having a 1 M transmission / reception structure has been established, and a 1 M transmission / reception structure and an IP network-based communication system have been established for providing a plurality of viewers with video packets in the field of video transmission. It is.

In a conventional communication system, a source may transmit a packet in real time by encoding the same image to a plurality of destinations, and the plurality of destinations may decode the packet to transmit or reproduce the image to a peripheral device. For example, assuming that the destination is Seoul, the same image can be transmitted to a plurality of domestic destinations such as Incheon, Daejeon, Gwangju, Daegu, and Busan, and the same image can be transmitted to a plurality of destinations outside Japan, China, the United States, and Europe. Can be.

Conventionally, in a 1M transmission / reception scheme, an image may be transmitted in a packet unit at a source, but packet loss may occur due to high quality or high capacity of the image and network load or congestion, and the image may be broken due to packet loss. There is a problem that the flow of the image can be interrupted, it is difficult to recover the lost packet. For example, a situation may occur where packets sent from a source are not normally delivered to a destination, and some of the packets are lost on a router.

Korean Patent Registration No. 10-1240895

In order to solve the above problems, the present invention provides N encoders that can be encoded into N video formats in consideration of packet loss, and video encoding for transmitting source packets to M destinations using an M × N transmission control matrix. An apparatus and a method of operating the same are provided.

The present invention provides an image encoding apparatus and an operation method thereof for changing component values of a transmission control matrix by dividing a state when an encoder switching request information is received from a destination where packet loss occurs and a state when the encoder completes switching. do.

The present invention allows one context switching and one copy of the source packet of the i-th encoder from user space to kernel space when j number of second component values are included in the column direction of the i-th encoder in the transmission control matrix. The present invention provides a video encoding apparatus for transmitting a source packet copied from kernel space j times to j destinations and a method of operating the same.

According to an aspect of the present invention, there is provided an apparatus for encoding an image using a transmission control matrix, including: an encoder including N encoders capable of encoding N image formats; A first generation unit generating selection division information for distinguishing a destination corresponding to each selected encoder; A MxN transmission control matrix corresponding to N encoders and M destinations is generated, and a first component value representing an encoder and a destination that is not selected in the transmission control matrix through selection classification information, and a second component representing the selected encoder and the destination. A second generation unit reflecting a value and a communication unit transmitting a source packet to a destination corresponding to the encoder selected through the transmission control matrix, with reference to the second component value representing the encoder and the destination selected in the transmission control matrix. The source packet is transmitted to M destinations.

When the first generation unit generates selection conversion information for switching from the i-th encoder selected corresponding to the j-th destination to the k-th encoder, the second generation unit corresponds to the j-th destination and the k-th encoder through the selection conversion information. The first component value may be changed to the third component value.

In an image encoding apparatus using a transmission control matrix according to an embodiment of the present invention, a queue for temporarily storing a source packet generated by the encoding unit and a source packet transmitted from the queue to the communication unit may be an IDR frame encoded by a k-th encoder. The apparatus may further include a monitoring unit that monitors the second generation unit, and when the IDR frame is detected, the second generation unit changes the third component value corresponding to the j-th destination and the k-th encoder to the second component value, and corresponds to the j-th destination and the i-th encoder. The second component value may be changed to the first component value.

According to an embodiment of the present invention, an operation method of an image encoding apparatus using a transmission control matrix including N encoders capable of encoding into N image formats may include generating selection classification information for distinguishing a corresponding destination for each selected encoder. ; A MxN transmission control matrix corresponding to N encoders and M destinations is generated, and a first component value representing an encoder and a destination that is not selected in the transmission control matrix through selection classification information, and a second component representing the selected encoder and the destination. Inputting a value and transmitting a source packet to a destination corresponding to the encoder selected through the transmission control matrix, wherein the M destinations are referenced by referring to a second component value representing the encoder and the destination selected in the transmission control matrix. The source packet is characterized in that for transmitting.

The transmitting of the source packet to the destination may include one context for the source packet of the i-th encoder from user space to kernel space when j number of second component values are included in the transmission control matrix in the column direction of the i-th encoder. Switching and one copy are made, and the source packet copied in kernel space is repeatedly transmitted j times to j destinations.

According to the present invention, by providing N encoders that can be encoded into N video formats, the encoder can be selected according to the network conditions for each destination, the packet loss rate for each destination can be reduced, and the packet loss rate for each destination is reduced, thereby preventing the video from being broken. Or to prevent the flow of the image.

According to the present invention, by changing the component value of the transmission control matrix, the image is not broken during the switching process of the encoder. Or to prevent interruption of the video flow.

According to the present invention, one context switching and one copy is performed from user space to kernel space when the same data is transmitted to a plurality of destinations, thereby reducing the load and power consumption of the video encoding apparatus.

1 illustrates an image encoding apparatus according to an embodiment of the present invention.
FIG. 2 is a detailed block diagram illustrating the encoding unit of FIG. 1.
3 shows an example of a transmission control matrix of M × N.
4 shows an example of a state change of a 3x3 transmission control matrix.
5 illustrates a method of operating an image encoding apparatus according to an embodiment of the present invention.
6 illustrates an example of context switching and copying from user space to kernel space.
7 shows an example of context switching and copying from user space to kernel space of the present invention.

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings and the contents described in the accompanying drawings, but the present invention is not limited or limited to the embodiments.

1 illustrates an image encoding apparatus according to an exemplary embodiment of the present invention. The image encoding apparatus 100 transmits source packets to M destinations 300 using a transmission control matrix. The image encoding apparatus 100 includes an encoding unit 110, a queue 120, a communication unit 130, and a controller 140.

The encoding unit 110 generates a source packet of an image stream by encoding the source image, and sends the source packet to the queue 120 one frame at a time. The source image may be an image received through an external terminal such as HDMI or SDI, or may be an image stored in a storage medium.

The queue 120 temporarily stores the source packet, and the communication unit 130 extracts the source packet temporarily stored in the queue 120 and transmits the source packet through the IP network 200. The controller 140 controls operations of the encoding unit 110, the queue 120, and the communication unit 130.

FIG. 2 is a detailed block diagram illustrating the encoding unit of FIG. 1, wherein the encoding unit 110 may encode N encoders 110-1, 110-2,. N). Each of the N encoders 110-1, 110-2,..., 110 -N supports different video formats such as codec characteristics, video quality, frames per second, and compression rate.

If the packet loss occurs when the source packet of the selected encoder is transmitted to the specific destination through the communication unit 130, the encoding unit 110 switches to another encoder having a reduced image quality, frames per second or compression rate than the selected encoder. As shown in FIG. 2, the N encoders 110-1, 110-2,..., And 110 -N are included. For example, the encoding unit 110 transmits the source packet of the first encoder 110-1 that supports the H.265 1080p, 60 fps, and 5 Mbps image formats to the first destination 300-1 through the communication unit 130. If packet loss occurs during transmission, the second encoder 110-2 supports a video format of 30 fps and 3 Mbps having a reduced frame rate and compression rate per second than the first encoder 110-1, and a second encoder. The source packet of 110-2 may be transmitted to the first destination 300-1 to reduce the packet loss probability.

The controller 140 includes a first generator 141, a second generator 142, and a monitor 143. The first generator 141 generates selection classification information for distinguishing a destination corresponding to each selected encoder.

3 illustrates an example of an M × N transmission control matrix. The second generation unit 142 generates an M × N transmission control matrix corresponding to the number of encoders and destinations. The second generation unit 142 is E ₁ , E ₂ ,. N encoder rows such as, E _N , and T ₁ , T ₂ ,... Create M destination columns, such as T _M.

The second generation unit 142 reflects the first component value indicating the non-selected encoder and the destination and the second component value indicating the selected encoder and the destination in the transmission control matrix through the selection division information.

When the first generator 141 generates selection switch information for switching from the i-th encoder selected to the k-th encoder corresponding to the j-th destination, the second generator 142 generates the j-th destination through the selection switch information. And the first component value corresponding to the kth encoder is changed to the third component value. When the communication unit 140 receives the switch request information for requesting the switch of the encoder from the j-th destination, the communication unit 140 transmits the switch request information to the first generator 141, and the first generator 141 transmits the selection switch information based on the switch request information. Create The monitoring unit 143 monitors whether the source packet transmitted from the queue 120 to the communication unit 130 is an IDR frame encoded by the k-th encoder. When the IDR frame is detected, the second generation unit changes the third component value corresponding to the j-th destination and the k-th encoder to the second component value, and simultaneously removes the second component value corresponding to the j-th destination and the i-th encoder. 1 Change to the component value.

The source packet includes a P frame, a B frame, and an IDR frame. The IDR frame is called an instantaneous decoding refresh frame and serves as a key frame. The P frame is called a previous or predicted frame and is a predicted frame based on information of a previous IDR frame. The B frame is called a bidirectional frame and is a frame of fore and aft association based on information of an IDR frame and a P frame. When the P 130 or B frames such as the correlation between the prediction frame and the front and rear frames are transmitted as the first frame, the communication unit 130 may cause an image to be broken. Therefore, the communication unit 130 transmits the IDR frame from the first frame of the source packet of the conversion encoder. In addition, the monitoring unit 143 monitors the IDR frame because it does not know what kind of frame is coming from the queue 120 to the communication unit 130.

The source packet may include information on which type of frame is a P frame, a B frame, or an IDR frame, and may include information on which encoder was generated. Information about any encoder can be added in the form of a tag.

In Figure 4 is an example showing a state change of the transmission control matrix of 3 × 3, the second generation part 142 is E _1, T ₂ and T ₃ corresponding to T ₁ via the selection indicator information to the transmission control matrix reflects the second component value representing a first component value and the rest of the component values of the selected encoder and the destination represents the destination and the non-selected encoder E ₃ and so on corresponding to the corresponding E ₂ and T ₃ to T ₁ ~. For example, the first component value may be '0' and the second component value may be '1'.

The communication unit 130 transmits a source packet to a destination corresponding to the selected encoder with reference to the second component value of the transmission control matrix. When the communication unit 130 receives the switching request information for requesting the switching of the encoder from T _{2, the} communication unit 130 transmits the switching request information to the first generating unit 141, and the first generating unit 141 generates the selection switching information based on the switching request information. do. T ₂ may transmit a switch request information to the communication unit 130 because a loss occurs in a portion of a previously received source packet.

The first generator 141 generates selection switch information for switching from E ₁ selected corresponding to T ₂ to E ₃ capable of reducing packet loss. The second generation unit 142 changes the first component value corresponding to T ₂ and E ₃ to the third component value through the selection change information. For example, the third component value may be '2'.

The monitoring unit 143 monitors whether the packet transmitted from the queue 120 to the communication unit 130 is an IDR frame encoded at E ₃ , and the second generating unit 142 transmits a signal to T ₂ and E ₃ when the IDR frame is detected. The corresponding third component value is changed to the second component value, and the second component value corresponding to T ₂ and E ₁ is changed to the first component value.

According to the present invention, by changing the component value of the transmission control matrix, the image is not broken during the switching process of the encoder. Or to prevent interruption of the video flow.

FIG. 5 illustrates a method of operating an image encoding apparatus according to an exemplary embodiment of the present invention. The image encoding apparatus 100 generates selection classification information for distinguishing a corresponding destination for each selected encoder, and generates selection classification information. A transmission control matrix of M × N is generated, and a source packet is transmitted to a destination corresponding to the selected encoder through the transmission control matrix.

FIG. 6 illustrates an example in which context switching and copying are performed from a conventional user space to a kernel space. When the conventional scheme is applied, the image encoding apparatus 100 may determine j numbers in the column direction of the i th encoder in the transmission control matrix. When the second component value is included, j context switching and j copying may be performed on the source packet of the i-th encoder from user space to kernel space. However, in the image encoding apparatus 100 to which the conventional method is applied, the context switching and the copying of the source packet may occur for each number of destinations, thereby causing a system overload or an increase in power consumption.

7 illustrates an example in which context switching and copying is performed from user space to kernel space of the present invention. In the video encoding apparatus 100 of the present invention, j number of items in the column direction of the i-th encoder in the transmission control matrix are illustrated. If two component values are included, one context switching and one copy is performed for the source packet of the i-th encoder from user space to kernel space, and the source packet copied in kernel space is repeatedly transmitted j times to j destinations. do.

The video encoding apparatus 100 of the present invention may call the destination transfer function in kernel space and transmit the source packets of the i-th encoder, which has been copied once, to j destinations. The destination transfer function can be sendto (), a Berkeley socket system call function.

The present invention can reduce the load and power consumption of the video encoding apparatus 100 by performing one context switching and one copy of a source packet.

100: video encoding device 110: encoding unit
120: queue 130: communication unit
140: control unit 141: first generation unit
142: second generation unit 143: monitoring unit
200: IP network 300: M destinations

Claims (5)

An encoding unit including N encoders capable of encoding the N image formats;
A first generation unit generating selection division information for distinguishing a destination corresponding to each selected encoder;
A MxN transmission control matrix corresponding to N encoders and M destinations is generated, and a first component value representing an encoder and a destination that is not selected in the transmission control matrix through selection classification information, and a second component representing the selected encoder and the destination. A second generator that reflects the value;
Including a communication unit for transmitting a source packet to a destination corresponding to the selected encoder through the transmission control matrix,
And a source packet to M destinations by referring to an encoder selected in the transmission control matrix and a second component value representing a destination. The method of claim 1,
When the first generation unit generates selection conversion information for switching from the i-th encoder selected corresponding to the j-th destination to the k-th encoder, the second generation unit corresponds to the j-th destination and the k-th encoder through the selection conversion information. An image encoding apparatus using a transmission control matrix, wherein the first component value is changed to a third component value. The method of claim 2,
A queue for temporarily storing a source packet generated by the encoding unit;
A monitoring unit for monitoring whether the source packet transmitted from the queue to the communication unit is an IDR frame encoded by the k-th encoder,
When the IDR frame is detected, the second generation unit changes the third component value corresponding to the j-th destination and the k-th encoder to the second component value, and changes the second component value corresponding to the j-th destination and the i-th encoder to the first component. An image encoding apparatus using a transmission control matrix, characterized in that the component value is changed. In the operating method of an image encoding apparatus comprising N encoders capable of encoding in N image formats,
Generating selection classification information for distinguishing a destination corresponding to each selected encoder;
A MxN transmission control matrix corresponding to N encoders and M destinations is generated, and a first component value representing an encoder and a destination that is not selected in the transmission control matrix through selection classification information, and a second component representing the selected encoder and the destination. Entering values and
Transmitting a source packet to a destination corresponding to the selected encoder through the transmission control matrix;
And transmitting source packets to M destinations by referring to an encoder selected in the transmission control matrix and a second component value representing a destination. The method of claim 4, wherein
The step of transmitting a source packet to the destination,
In the transmission control matrix, if j number of second component values are included in the column direction of the i-th encoder, one context switching and one copy is performed for the source packet of the i-th encoder from user space to kernel space. A method of operating an image encoding apparatus using a transmission control matrix, characterized by repeatedly transmitting a source packet copied from a space to j destinations repeatedly.

Images