JPH1032827A - Animation image encoder and animation image decoding system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To prevent image quality deterioration which is caused by underflow restriction of transmitting and receiving buffer in a device which encodes animation image signals. SOLUTION: An offset is operated and stored in an offset storing means 52 by virtual buffer occupancy amount which is at the head of a GOP(group of pictures) and is stored in a virtual buffer occupancy amount storing means 23. In a frame target operating means 21, a frame target in each frame in operated by distributing the value which subtracts the offset from GOP residual amount, stored in a GOP residual amount storing means to each frame in the GOP. Buffer occupancy amount at the head of the GOP is increased by suppressing the bit generating amount of the GOP by the buffer occupancy amount at the head of the GOP, and image quality deterioration which is caused by underflow restriction in the buffer can be prevented.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a moving picture coding apparatus for coding a video signal, and a method for transmitting a coded moving picture on a transmission line or storing the coded moving picture in a storage medium and decoding the coded moving picture into an original video signal. The present invention relates to a moving picture decoding system to be transformed.

[0002]

2. Description of the Related Art MPEG2 (ISO / IEC 1381)
In encoding of a moving image such as 8-2), the amount of generated data differs between frames. When such data is transmitted over a transmission path with a constant bit rate, the video encoding device controls the amount of data generated in the frame so that overflow or underflow does not occur in the reception buffer of the decoding device on the receiving side. There is a need to. Also, when the encoded data is stored in a storage medium, and then the stored data is read out and decoded by the decoding device, it is necessary to control the data generation amount by the moving image coding device.

As an example of such a control method, INTERNAT
IONAL ORGANISATION FOR STANDARDISATION, CEDED REPR
ESENTATION OF PICTURE AND AUDIO INFORMATION, Test
Model 3 (ISO-IEC / JTC1 / SC29 / WG11). Hereinafter, a conventional moving picture coding apparatus will be described.

Each frame input as a moving image has I
There is a frame type of any of a frame, a P frame, and a B frame. These frame types are repeated on a group of picture (GOP) basis with the I frame period being N and the I or P frame period being M.
Here, a case of a GOP with N = 15 and M = 3 will be described. The structure of the GOP in this case is as follows. IBBPBBPBPBPBPBP
The BB video coding apparatus performs coding in a coding mode determined for each frame type, and outputs the result as a bit stream. At this time, it is necessary to control so that the data generation amount of each GOP becomes equal and the reception buffer of the decoding device does not fail. Therefore, the moving picture coding apparatus assumes a reception buffer of the decoding apparatus at the time of coding, sets this as a virtual buffer, and performs coding so that the buffer does not fail.

[0005] After the conventional moving picture coding apparatus is described below, the frame target calculation unit of the moving picture coding apparatus will be described. First, a conventional moving picture coding apparatus will be described with reference to FIG. As an example, the video signal is IT
U recommendation 601 4: 2: 0. The number of pixels in one frame is 720 for the luminance signal (hereinafter referred to as Y signal) in the horizontal direction, 480 for the vertical direction, and the color difference R signal (hereinafter referred to as P signal).
r signal) and color difference B signal (hereinafter referred to as Pb signal)
Is 360 pixels in the horizontal direction and 240 pixels in the vertical direction. In this video signal, the Y signal is a block of 16 pixels in the vertical direction and 16 pixels in the horizontal direction.
The signal b is input with a block of 8 pixels in the vertical direction and 8 pixels in the horizontal direction as one unit. The block of the Y signal, the block of the Pr signal, and the block of the Pb signal are combined into a macro block. One frame is composed of 1350 macroblocks.

[0006] In FIG. 10, a frame type input terminal 1 inputs a frame type of I, P, B. A video signal is input to the video signal input terminal 2. The encoding means 3 encodes the video digital signal in macroblock units in the encoding mode specified by the frame type, and outputs the encoded bit stream to a bit stream output terminal 4.
Output more. At this time, by controlling the quantization step used for quantization with the quantization coefficient, the bit generation amount is controlled in macroblock units. That is, by increasing the quantization coefficient, the quantization step is increased, and the amount of macroblock bits generated (hereinafter, macroblock bit generation amount) is suppressed. Conversely, by reducing the quantization coefficient, the quantization step becomes finer and the amount of macroblock bit generation increases. The macroblock bit generation amount counting means 5 measures the macroblock bit generation amount of the bit stream. The frame bit generation amount counting means 6 measures the bit generation amount (hereinafter, referred to as frame bit generation amount) for each frame of the bit stream by adding the macroblock bit generation amount for one frame. The quantization coefficient addition means 7 measures the total sum of the quantization coefficients in one frame.
The frame target calculation unit 8 calculates a frame target, which is a target frame bit generation amount of the frame, from the frame bit generation amount of the already encoded frame and the sum of the frames of the quantization coefficients used at that time. Is what you do. The quantization coefficient calculation means 9 calculates the quantization coefficient for each macroblock, and increases or decreases the quantization coefficient so that the frame bit generation amount matches the frame target.

[0007] The operation of the moving picture coding apparatus configured as described above will be described. At the beginning of encoding a frame,
From the frame type input from the frame type input terminal 1, the sum of the quantized coefficients counted by the quantized coefficient adding means 7 and the frame bit generation counting means 6 and the frame bit generation amount, respectively, the frame target calculating section 8
To calculate the frame target. After calculating the frame target, for one frame per macroblock,
The following operation is repeated to perform encoding.

Before coding the macro block, the quantization coefficient calculating means 9 calculates the macro block bit generation amount measured by the macro block bit generation amount counting means 5 and the frame calculated by the frame target calculating section 8. The quantization coefficient is calculated from the target. Then, the encoding means 3 encodes the frame type input from the frame type input terminal 1 and the macroblock of the video signal input from the video signal input terminal 2, and outputs a bit stream from the bit stream output terminal 4. . At this time, by the quantization coefficient calculated by the quantization coefficient calculation means 9,
The bit generation amount is controlled. The bit generation amount of the output bit stream is measured by the macroblock bit generation amount counting means 5.

After the encoding of one frame is completed, the frame bit generation amount counting means 6 measures the frame bit generation amount from the macroblock bit generation amount of the macroblock bit generation amount counting means 5, and adds the quantization coefficient addition means. Numeral 7 measures the sum of the quantized coefficients of the quantized coefficient calculating means 9.

Next, the frame target calculating section 8 of the conventional moving picture coding apparatus will be described with reference to FIG. FIG.
, The sum of the quantized coefficients output from the quantized coefficient adding means 7 in FIG. 10 is input to the quantized coefficient input terminal 11. As shown in FIG.
The frame bit generation amount output from the 0 frame bit generation amount counting means 6 is input. The frame type input terminal 13 shown in FIG.
The input frame type is input. After the encoding of one frame is completed, the complexity calculating means 14 calculates the product of the sum of the quantized coefficients and the amount of generated frame bits (hereinafter, this product is referred to as the complexity). The complexity is constant irrespective of the quantization coefficient for the same image, and its value increases on a screen with a complicated image, and thus serves as an index indicating the complexity of the image. The switching means 15 outputs the inputted complexity to the complexity storage means 16 for an I frame, to the complexity storage means 17 for a P frame, and to the complexity storage means 18 for a B frame. The complexity storage means 16, 17, and 18 store the complexity of I, P, and B frames, respectively.

A target value of a bit generation amount generated in the GOP is defined in one GOP, and a bit generation target value (hereinafter, referred to as a GOP target) TR is set at the start of encoding. The GOP remaining amount calculating means 19 sets the GOP target as the GOP remaining amount before encoding the first frame, and updates the GOP remaining amount by subtracting the frame bit generation amount from the GOP remaining amount after the second frame. And 1 GO
After the encoding of P is completed, the GOP
The GOP remaining amount is obtained by adding the target TR. The GOP remaining amount storage means 20 stores the GOP remaining amount thus calculated. Before encoding the frame, the frame target computing means 21 distributes the GOP remaining amount based on the following equation according to the complexity of each frame type already encoded and the frame type to be encoded. Calculate the target (1). For example, G
The remaining number of OPs is Rn, the numbers of I-frames, P-frames, and B-frames that have not been encoded in the current GOP are Ni, Np, and Nb, respectively. When the complexity of the frame and the complexity of the B frame are Xi, Xp, and Xb, respectively,
The frame targets (1) Ti, Tp, and Tb of the I frame, the P frame, and the B frame are as follows. Ti = R / Ai Ai = 1 + Np * Xp / Xi + Nb * Xb / (1.4 *
Xi) Tp = R / Ap Ap = Np + Nb * Xb / Xp Tb = R / Ab Ab = Nb + 1.4 * Np * Xp / Xb

The virtual buffer occupancy calculating means 22 updates the initial value set before coding or the occupancy of the virtual buffer of the immediately preceding frame for each frame. The virtual buffer occupancy is calculated by subtracting the frame bit generation amount from the virtual buffer occupancy after the frame is encoded, and adding a value obtained by dividing the GOP target by the number of GOP frames. The virtual buffer occupancy storage means 23 stores the calculated virtual buffer occupancy. The underflow detection means 24 sets the virtual buffer occupancy as the frame target (2) when the frame target is larger than the virtual buffer occupancy, and otherwise sets the frame target (1) as the frame target (2). The overflow detecting means 25 adds the virtual buffer occupancy to the value obtained by dividing the GOP target by the number of frames of the GOP, and then reduces the capacity of the virtual buffer (hereinafter, this value is referred to as the lower limit of the target) and the frame target. Compare (2). When the frame target (2) is smaller than the lower limit of the target, the lower limit of the target is set as the frame target (3). Otherwise, the frame target (2) is set as the frame target (3). The frame target output terminal 26 is a terminal for outputting a frame target.

The operation of the conventional frame target calculation unit 8 configured as described above is shown in FIGS.
This will be described with reference to FIG. FIG. 12 shows a frame type input from the frame type input terminal 12 during one GOP (15 frame periods), and a GOP at the start of encoding of each frame.
The GOP remaining amount (R
n), the virtual buffer occupancy (Vn) stored in the virtual buffer occupancy storage means 23, the I-frame complexity (XIn) stored in the complexity storage means 16, and the complexity storage means 17 , The complexity of the P frame stored in the complexity storage unit 18, the complexity of the B frame stored in the complexity storage unit 18, and the frame target output from the frame target calculation unit 21.
(Tn) and the bit generation amount (Gn) of the frame. n indicates a frame number (1 to 15).

FIG. 13 shows a processing procedure of each of the following steps performed by the frame target calculating section 8. Step 1: Setting the remaining GOP At the beginning of the GOP, the GOP remaining
The GOP target (TR) is added to the GOP remaining amount of the last frame of the previous GOP stored in the OP remaining amount storage unit 11.
Add. Then, the value R1 is stored in the GOP remaining amount storage unit 1.
Set to 1.

Step 2: Calculation of frame target Before encoding the frame of the first frame, the frame target calculation means 21 calculates the complexity (I) of the I frame stored in the complexity storage means (1) 16. XI1), complexity storage means (2)
17, the complexity (XP1) of the P frame and the complexity (XB) of the B frame stored in the complexity storage means (3) 18.
1), frame type (I) inputted from the frame type input terminal (2) 12, and GOP remaining amount storage means 1
The frame target (1) T1 is calculated from the GOP remaining amount (R1) stored in No. 1.

Next, the virtual buffer occupancy (V1) stored in the virtual buffer occupancy storage means 23 by the underflow detection means 24 and the frame target calculation means 21
The frame output T1 is compared with the output frame target T1.
When V1> T1, no underflow occurs as shown in FIG. 15A, so T1 is output as the frame target T2. When V1 ≦ T1, underflow occurs, so V1 is output as the frame target T2. The overflow detecting means 25 adds the virtual buffer occupancy (V1) stored in the virtual buffer occupancy storage means 23 to a value obtained by dividing the GOP target (TR) by the number of frames of the GOP (N = 15). Lower limit value of the target obtained by subtracting the capacity (Vfull) of the virtual buffer from the value (Tlower1 = TR / N + V1-Vfull)
And the frame target T2 output from the underflow detection means 24. T2> Tlower1
At this time, even if the frame bit generation amount of the frame to be encoded is 0, the overflow does not occur.
T2 is output to the frame target output terminal 26 as T2 as the frame target T3, and Tlower1 at other times to avoid overflow.

Step 3 Next, one frame is encoded by the frame encoding means 3.

Step 4: Complexity Calculation After encoding the first frame, the complexity calculating means 24 inputs the quantization coefficient input from the quantization coefficient input terminal 11 and the frame bit generation amount input. Terminal 1
The complexity XI2 is calculated from the frame bit generation amount inputted from the frame type 2 and the switching means 15 stores the complexity storage means (1) into the XI2 from the frame type (I) inputted from the frame type input terminal 12. Update.

Step 5: Calculation of remaining GOP In the GOP remaining amount calculating means 19, the frame input from the frame bit generation amount input terminal 12 is obtained from the GOP remaining amount (R1) stored in the GOP remaining amount storage means 20. The bit generation amount G1 is subtracted, and its value (R2 = R1-G1)
Then, the GOP remaining amount storage means 20 is updated.

Step 6: Calculation of virtual buffer occupancy The virtual buffer occupancy calculation means 22 calculates the frame bit generation amount input terminal 12 from the virtual buffer occupancy (V1) stored in the virtual buffer occupancy storage means 23.
To the value obtained by subtracting the input frame bit generation amount G1 from the GOP target (TR) by the number of frames of the GOP (N = 15), and adding the value (V2 = V1-
G1 + TR / N), and the virtual buffer occupancy storage means 23
To update.

The operations from step 2 to step 6 are repeated until the 15th frame. After the second frame, the operation is the same as that of the first frame, except for the operation of the switching means 15 and the arithmetic expression used in the frame target arithmetic means 21, and a description thereof will be omitted.

The following can be said by controlling the bit generation amount as described above. 1) The occupancy of the virtual buffer changes within the capacity of the virtual buffer without overflow or underflow. 2) The bit generation amount of the GOP is almost equal to the GOP target (T
R).

FIG. 14 shows a moving picture coding / decoding system using the above-mentioned conventional moving picture coding apparatus. The moving picture coding apparatus 10 is the apparatus shown in FIG. The transmission path 32 transmits the bit stream at a constant rate. The reception buffer 33 stores the input bit stream at a constant rate, and outputs the bit stream in response to a request from the video decoding device 34.

The bit stream output from the video encoder 10 is input to the transmission buffer 31 and then output to the transmission line 32 as a bit stream read at a constant bit rate. The bit stream having a constant bit rate transmitted through the transmission path 32 is input to the reception buffer 33 and then decoded by the video decoding device 34. In this case, the GOP target (TR [bit])
Indicates that the number of frames of the GOP is N, the bit rate of the transmission path is C [bps], and the frame frequency of the video signal is Fr [H
z], TR = C * N / Fr. By setting the virtual buffer capacity Vfull to a value smaller than the buffer capacity of either the transmission buffer 31 or the reception buffer 33, the transmission buffer 31
Both the receiver and the receiving buffer 33 can transmit a bit stream at a constant bit rate C without underflow or overflow.

[0025]

However, in the conventional moving picture coding apparatus, when a bit stream is output via a transmission buffer and a transmission path having a constant bit rate, the bit generation amount of each GOP and the transmission path Are controlled so that the bit amount of the bit stream output to the GOP becomes the same, the virtual buffer occupancy at the head of the GOP becomes substantially the same for each GOP. In the moving picture compression method of MPEG2, I-frames for which intra-frame encoding is performed are assigned more frame bits than P and B frames which perform both inter-frame encoding and intra-frame encoding. Then, at the beginning of the GOP, if the value of the occupation amount of the virtual buffer is small, the underflow is restricted, and encoding is performed with a frame bit amount smaller than the assigned target, the image quality of the GOP deteriorates. Since this phenomenon is controlled so that the number of bits transmitted in one GOP period and the bit generation amount of the GOP become the same, the following G
It also occurs repeatedly in OP. Further, when the image quality of a specific frame is degraded in the GOP, the uniformity of the image quality in the GOP is lost, and thus there is a problem that it is not preferable in terms of visual characteristics.

An object of the present invention is to solve the above-mentioned conventional problems, and an object of the present invention is to provide a moving picture coding apparatus and a moving picture decoding system which prevent deterioration of picture quality due to limitation of virtual buffer occupancy.

[0027]

According to a first aspect of the present invention, there is provided a method for inputting a moving image so that a bit generation amount of a GOP target is obtained for each GOP in which an I frame, a P frame, and a B frame are repeated in a predetermined order. Encoding means for encoding a signal according to any of the input frame types; frame bit generation amount counting means for counting the frame bit generation amount of a moving image encoded by the encoding means; From the frame bit generation amount counted by the bit generation amount counting means, G
A frame target operation unit for updating the remaining OP number and outputting a target bit generation amount of the frame according to the input frame type; and the encoding unit according to a frame target output from the frame target operation unit And a quantizing coefficient calculating means for generating a quantizing coefficient for encoding in the video encoding apparatus, wherein the frame target calculating section generates frame bits for each frame from a target GOP target. GOP remaining amount storage means for storing the GOP remaining amount updated based on the amount;
Offset calculating means for calculating an offset which is the difference between the virtual buffer occupancy at the beginning of each GOP and the virtual buffer occupancy target value; offset storing means for storing the offset calculated by the offset calculating means; Subtraction means for correcting the GOP remaining amount by subtracting the offset from the GOP remaining amount held in the remaining amount storage means;
And a frame target calculating means for calculating a target of a bit generation amount in a frame based on the GOP remaining amount corrected by the subtracting means.

[0028] In the invention of claim 2 of the present application, the moving picture coding apparatus inputs a quantized coefficient from the coding means,
And a quantizing coefficient adding unit that outputs a sum of the quantized coefficients of one frame after completing encoding of one frame, wherein the encoding unit includes a quantizing unit; The frame target calculating means includes: a GOP remaining amount stored in the GOP remaining amount storage means;
A frame target is calculated from the offset stored in the offset storage means and the sum of the quantization coefficients of one frame of the quantization coefficient addition means.

According to the invention of claim 3 of the present application, the moving picture coding apparatus detects underflow of a virtual buffer from a frame target output from the frame target calculating means and a virtual buffer occupancy, and detects underflow. The apparatus further comprises an underflow detecting means for subtracting the frame target when it is performed and outputting the result to the encoding means.

[0030] In the invention of claim 4 of the present application, the moving picture coding apparatus detects an overflow of the virtual buffer from the frame target and the virtual buffer occupancy output from the underflow detecting means, and detects the overflow. The apparatus further comprises overflow detecting means for increasing the number of frame targets and outputting the result to the encoding means.

According to a fifth aspect of the present invention, when the video encoding device detects an overflow of the virtual buffer from the virtual buffer occupancy, the video encoding device corrects the virtual buffer occupancy to the upper limit value of the virtual buffer. It is characterized by further comprising means.

According to a sixth aspect of the present invention, there is provided the moving picture encoding apparatus according to the second aspect, an output buffer for inputting a bit stream output from the moving picture encoding apparatus, and an output buffer input to the output buffer. A transmission path for transmitting the bit stream, a reception buffer for inputting the bit stream transmitted on the transmission path, and a video decoding device for decoding the bit stream input to the reception buffer and outputting a video signal And characterized in that:

According to a seventh aspect of the present invention, there is provided the moving picture coding apparatus according to the third aspect, an output buffer for inputting a bit stream output from the moving picture coding apparatus, and an output buffer input to the output buffer. A transmission path for transmitting the bit stream, a reception buffer for inputting the bit stream transmitted on the transmission path, and a video decoding device for decoding the bit stream input to the reception buffer and outputting a video signal And characterized in that:

According to an eighth aspect of the present invention, there is provided the moving picture coding apparatus according to the fourth aspect, an output buffer for inputting a bit stream output from the moving picture coding apparatus, and an output buffer input to the output buffer. A transmission path for transmitting the bit stream, a reception buffer for inputting the bit stream transmitted on the transmission path, and a video decoding device for decoding the bit stream input to the reception buffer and outputting a video signal And characterized in that:

According to a ninth aspect of the present invention, there is provided the moving picture coding apparatus according to the fifth aspect, an output buffer for inputting a bit stream output from the moving picture coding apparatus, and an output buffer input to the output buffer. A transmission path for transmitting the bit stream, a reception buffer for inputting the bit stream transmitted on the transmission path, and a video decoding device for decoding the bit stream input to the reception buffer and outputting a video signal And characterized in that:

According to a tenth aspect of the present invention, there is provided the moving picture encoding apparatus according to the second aspect, recording means for recording a bit stream output from the moving picture encoding apparatus, and recording by the recording means. A moving image decoding device that decodes the bit stream and outputs a video signal.

According to an eleventh aspect of the present invention, there is provided the moving picture coding apparatus according to the third aspect, recording means for recording a bit stream output from the moving picture coding apparatus, and recording by the recording means. A moving image decoding device that decodes the bit stream and outputs a video signal.

According to a twelfth aspect of the present invention, there is provided the moving picture encoding apparatus according to the fourth aspect, recording means for recording a bit stream output from the moving picture encoding apparatus, and recording by the recording means. A moving image decoding device that decodes the bit stream and outputs a video signal.

According to a thirteenth aspect of the present invention, there is provided the moving picture encoding apparatus according to the fifth aspect, recording means for recording a bit stream output from the moving picture encoding apparatus, and recording by the recording means. A moving image decoding device that decodes the bit stream and outputs a video signal.

Claims 1 to 1 of the present application having such features.
According to the fifth aspect of the present invention, the frame target operation unit is GOP
The amount of generated frame bits is reduced for each frame from the target, and the remaining GOP is updated and stored. And GOP
The virtual buffer occupancy target value is determined in advance at the beginning, and the difference from the virtual buffer occupancy at that time is calculated as an offset and stored in the offset storage means.
The GOP remaining amount is corrected by subtracting the offset from the GOP remaining amount, and the target of the bit generation amount of the frame is calculated based on the corrected GOP remaining amount.
For this reason, the virtual buffer occupancy at the head of the GOP can be made closer to the virtual buffer occupancy target value, and deterioration of the image quality of the I frame at the head of the GOP can be prevented.

The invention according to claims 6 to 9 is for transmitting and decoding a moving picture via a transmission path using the moving picture coding apparatus. The bit stream encoded by the image encoding device is recorded,
This is a moving image decoding system that decodes a recorded bit stream to reproduce a moving image.

[0042]

DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, a moving picture coding apparatus according to a first embodiment of the present invention will be described. The description will be given assuming that the input video signal is the same as that used in the description of the conventional moving picture coding apparatus. First, a video encoding device according to the first embodiment of the present invention will be described with reference to FIG. Note that the moving picture coding apparatus according to the first embodiment of the present invention shown in FIG. 2 has basically the same configuration as the conventional moving picture coding apparatus shown in FIG. The same numbers are assigned and the detailed description is omitted. The frame target calculating section 41 calculates a frame target from the amount of generated frame bits of the already coded frame and the sum of the frames of the quantized coefficients used at that time.

Next, referring to FIG. 1, the frame target calculating section 4 of the moving picture coding apparatus according to the first embodiment of the present invention will be described.
1 will be described. Note that the frame target calculation unit 41 of the moving picture coding apparatus according to the first embodiment of the present invention shown in FIG.
Has basically the same configuration as that of the frame target calculation unit 8 of the conventional video encoding device shown in FIG. 10, and therefore, the same components will be assigned the same reference numerals and detailed description thereof will be omitted.
When the frame type of the I frame is input, the offset calculating means 51 calculates a value obtained by subtracting the virtual buffer occupancy from an arbitrary virtual buffer occupancy target value (TV) (hereinafter referred to as an offset). The virtual buffer occupancy target value TV is an arbitrary value close to the virtual buffer capacity Vfull. This target value is the capacity occupied by the virtual buffer after the I frame encoding process is completed. The offset storage means 52 stores the offset, and the subtraction means 53 subtracts the offset from the remaining GOP.

The operation of the frame target calculating section 41 of the moving picture coding apparatus according to the first embodiment of the present invention configured as described above will be described with reference to FIGS. FIG. 3 shows the GOP remaining amount (R) stored in the GOP remaining amount storage unit 20 at the start of encoding of each frame of 1 GOP (15 frame periods) input from the frame type input terminal 13, The virtual buffer occupancy (V) stored in the buffer occupancy storage means 23
n), the complexity (XIn) of the I frame stored in the complexity storage means 16, the complexity (XPn) of the P frame stored in the complexity storage means 17, and the complexity storage means 1
8, the complexity of the B frame (X
Bn), the frame target (1) (Tn) output from the frame target calculating means 21 and the bit generation amount (Gn) of each frame. n indicates a frame number (1 to 15).

FIG. 4 shows a processing procedure of each of the following steps performed by the frame target calculating section 41. Step 1: Setting the remaining GOP At the beginning of the GOP, the GOP remaining
The GOP target (TR) is added to the GOP remaining amount of the last frame of the previous GOP stored in the OP remaining amount storage unit 11.
Add. Then, the value R1 is stored in the GOP remaining amount storage unit 1.
Set to 1.

Step 2 In the second step, it is first determined whether or not the input video signal is an I frame. If it is an I frame, step 3,
For other frames, go to step 4.

Step 3: Offset calculation Since the first frame is an I frame, the offset calculation means 51 calculates the virtual buffer occupancy V1 stored in the virtual buffer occupancy storage means 23 from the virtual buffer occupancy target value (TV). Offset value (TV-V
1) is set in the offset storage means 52. This offset value is not updated during one GOP.

Step 4: Correction of the remaining GOP In the subtraction means 53, the offset (TV-V1) stored in the offset storage means 52 is calculated based on the remaining GOP (R1) stored in the remaining GOP storage means 20. Reduce.

Step 5: Calculation of Frame Target In the frame target calculation means 21, the complexity (XI1) of the I frame stored in the complexity storage means 16 and the complexity storage means 1
7, the complexity of the P frame (X
P1), the complexity (XB1) of the B frame stored in the complexity storage means 18, the frame type (I) input from the frame type input terminal (2) 13, and the subtraction by the subtraction means 53. The frame target T1 is calculated from the remaining GOP (R1- (TV-V1)). Next, the underflow detection means 24 compares the virtual buffer occupancy (V1) stored in the virtual buffer occupancy storage means 23 with the frame target T1 output from the frame target calculation means 21. When V1> T1, no underflow occurs, so
T1 is output as the frame target T2, and otherwise V1 is output because of underflow. Then, the value obtained by adding the virtual buffer occupancy (V1) stored in the virtual buffer occupancy storage means 23 to the value obtained by dividing the GOP target (TR) by the number of frames of the GOP (N = 15) by the overflow detection means 25. Is the capacity of the virtual buffer (V
lower) (Tlower)
1 = TR / N + V1-Vfull) and the frame target T2 output from the underflow detection means 24. When T2> Tlower1, overflow does not occur even if the amount of frame bit generation of a frame to be encoded is 0, so T2 is set as the frame target T3, and Tlower1 is set as the frame target output terminal otherwise because overflow occurs. 26.

Step 6 Next, the encoding means 3 encodes one frame.

Step 7: Calculation of Complexity After encoding the first frame, the complexity calculation means 14 inputs the quantization coefficient input from the quantization coefficient input terminal 11 and the frame bit generation amount input. The complexity XI2 is calculated from the frame bit generation amount input from the terminal 12, and the switching means 15 updates the complexity storage means 16 to XI2 from the frame type (I) input from the frame type input terminal 12. I do.

Step 8: Calculation of the remaining GOP In the GOP remaining calculation means 19, the frame input from the frame bit generation input terminal 12 is obtained from the GOP remaining (R1) stored in the GOP remaining storage means 20. The bit generation amount G1 is subtracted, and its value (R2 = R1-G1)
Then, the GOP remaining amount storage means 20 is updated.

Step 9: Calculation of virtual buffer occupancy The virtual buffer occupancy calculation means 22 calculates the frame bit generation amount input terminal 12 from the virtual buffer occupancy (V1) stored in the virtual buffer occupancy storage means 23.
To the value obtained by subtracting the input frame bit generation amount G1 from the GOP target (TR) by the number of frames of the GOP (N = 15), and adding the value (V2 = V1-
G1 + TR / N), the virtual buffer occupancy storage means storage 23 is updated.

After the second frame, the operations of steps 4 to 9 are repeated until the fifteenth frame.
After the second frame, the operation is the same as that of the first frame, except for the operation of the switching means 15 and the arithmetic expression used in the frame target arithmetic means 21, and a description thereof will be omitted.

As described above, the moving picture coding apparatus according to the first embodiment of the present invention obtains an offset at the head of a GOP in accordance with the virtual buffer occupancy by the offset calculating means, and then converts the GOP. The frame target computing means controls the encoding so as to perform encoding with the bit generation amount obtained by subtracting the offset from the GOP target. So the next GO
The virtual buffer occupancy at the beginning of P can be made closer to the virtual buffer occupancy target value. Therefore, by setting the virtual buffer occupancy target value (TV) to a value close to Vfull, the first I frame of the GOP is less likely to be restricted by underflow, and the image quality of the I frame is degraded and the image quality of the GOP is accordingly reduced. The effect of preventing the loss of uniformity can be obtained.

Hereinafter, a moving picture coding apparatus according to a second embodiment of the present invention will be described. The moving picture coding apparatus according to the second embodiment of the present invention relates to a case where a storage medium for storing moving picture data is used. In this case, except for the frame target calculation unit, FIG.
Since the configuration is the same as that of the moving picture coding apparatus according to the first embodiment of the present invention shown in FIG. 5, only the frame target calculation unit will be described here with reference to FIG. It should be noted that the frame target calculation unit 41A of the moving picture coding apparatus according to the second embodiment of the present invention shown in FIG. 5 basically includes the frame target calculation unit according to the first embodiment of the present invention shown in FIG. Since the configuration is the same as that of the unit 41, the same components will be assigned the same reference numerals and detailed description will be omitted. Also, the input video signal will be described as being the same as that used in the description of the conventional moving picture coding apparatus.

The overflow detecting means 55 inputs the virtual buffer occupancy, and when the overflow of the virtual buffer is detected, corrects the virtual buffer occupancy to the value of the capacity of the virtual buffer. This simulates a state in which reading of data from the storage medium is temporarily stopped in the moving picture decoding apparatus.

The frame target computing section 41A according to the second embodiment of the present invention having the above-described configuration will now be described.
The operation will be described below with reference to FIGS. FIG. 3 has already been described in the description of the operation of the video encoding device according to the first embodiment of the present invention, and a description thereof will be omitted.
FIG. 6 shows a processing procedure of each step of the frame target calculation unit 41A according to the second embodiment of the present invention. Except for the calculation of the frame target in step 5 and the detection of the overflow in step 11, the frame target calculation unit 4 according to the first embodiment of the present invention shown in FIG.
1 is basically the same as the processing procedure performed in step 1, and the description is omitted.

Step 5: Calculation of frame target In the frame target calculation means 21, the complexity (XI1) of the I frame stored in the complexity storage means 16 and the complexity storage means 1
7, the complexity of the P frame (X
P1), the complexity (XB1) of the B frame stored in the complexity storage means 18, the frame type (I) input from the frame type input terminal 12, and the GOP residue subtracted by the subtraction means 53. The frame target (1) T1 is calculated from the quantity (R1- (TV-V1)). The underflow detection unit 24 compares the virtual buffer occupancy (V1) stored in the virtual buffer occupancy storage unit 23 with the frame target (1) T1 output from the frame target calculation unit 21, and V1> T1 At this time, T1 is output to the frame target output terminal 26 as T1 as the frame target T2, and V1 is output at other times.

Step 11: Overflow Detection The overflow detection means 55 detects the overflow of the virtual buffer from the virtual buffer occupancy stored in the virtual buffer occupancy detection means 23, and when the overflow is detected, the virtual buffer occupancy To the value of the capacity of the virtual buffer.

As described above, the moving picture coding apparatus according to the second embodiment of the present invention calculates the offset at the head of the GOP according to the virtual buffer occupancy by the offset calculating means, and then converts the GOP. The frame target calculating means controls the coding so that the bit generation amount is obtained by subtracting the offset from the GOP target, so that the virtual buffer occupancy at the beginning of the next GOP is close to the virtual buffer occupancy target value. On the other hand, when the overflow of the virtual buffer is detected by the overflow detecting means, the virtual buffer occupancy is corrected to the value of the capacity of the virtual buffer. This allows
Even in a moving picture coding apparatus which is not restricted by overflow, the first I frame of the GOP is less likely to be restricted by underflow, so that the image quality of the I frame is deteriorated and the uniformity of the GOP image quality is lost. To avoid being disturbed.

Although the embodiment of the present invention has been described in connection with the case where the bit rate outputted via the buffer is constant, the case where the bit rate changes in GOP units can also be controlled by changing the GOP target. It has a similar effect. Also, the case has been described where the GOP structure has an I-frame cycle of 15, and the I or P-frame cycle has 3, but other GOP structures have the same effect. Furthermore, although MPEG2 is used as the encoding control method, other encoding methods such as determining the bit generation amount in a plurality of frames and distributing the bit generation amount in the frame units to determine the bit generation amount in each frame. The control system has the same effect.

Hereinafter, a moving picture decoding system according to a third embodiment of the present invention will be described with reference to FIG. FIG.
The video signal output device 71 outputs a video signal, and the video encoding device 72 according to the first embodiment of the present invention.
Encodes the video signal into a bit stream, and the transmission buffer 73 is a buffer whose capacity is larger than the capacity of the virtual buffer, which inputs the bit stream, converts it to the transmission rate of the transmission path, and outputs it. The transmission path 74 transmits the bit stream, and the reception buffer 75 is a buffer whose capacity is equal to or larger than the capacity of the virtual buffer. The video decoding device 76 decodes the bit stream into a video signal, and the video signal display device 77 displays the video signal.

The operation of the moving picture decoding system according to the third embodiment of the present invention configured as described above will be described below with reference to FIG. The video signal output from the video signal output device 71 is coded into a bit stream by the video coding device 72, and then output to the transmission path 74 via the transmission buffer 73. The bit stream transmitted on the transmission path 74 is decoded into a video signal by a video decoding device 76 via a reception buffer 75, and the video signal is displayed on the video signal display device 7.
7 is displayed.

As described above, the moving picture decoding system according to the third embodiment of the present invention uses the coding apparatus according to the first embodiment of the present invention to reduce the deterioration of the I-frame image quality. , Thereby preventing loss of uniformity of the image quality of the GOP.

Here, the coding apparatus according to the first embodiment of the present invention is used as the moving picture coding apparatus, but the moving picture coding apparatus according to the second embodiment of the present invention is used. Has the same effect.

Next, a video decoding system according to a fourth embodiment of the present invention will be described with reference to FIGS.
In FIG. 8, the video signal output device 81 outputs a video signal,
The moving picture coding apparatus 82 of the embodiment codes a video signal into a bit stream. The output buffer 83 is a buffer whose capacity is equal to or larger than the capacity of the virtual buffer, converts the rate of the input bit stream and outputs the converted bit stream, and the recording means 84 records the bit stream. FIG. 9 shows a portion for decoding a moving image from the medium of the recording means 91 which records the moving image. The bit stream is converted into a rate used by the moving image encoding device 93 via an input buffer 92 and output. I do. The input buffer 92 is a buffer whose capacity is equal to or larger than the capacity of the virtual buffer. The moving picture decoding device 93 decodes a bit stream into a video signal, and the video signal display device 94 displays the video signal.

The operation of the moving picture decoding system according to the fourth embodiment of the present invention configured as described above will be described with reference to FIGS. In FIG. 8, a video signal output from a video signal output device 81 is encoded into a bit stream by a video encoding device 82 according to the first embodiment of the present invention, and is recorded via an output buffer 83. Output to the means 84. The bit stream recorded in the recording unit 91 in FIG. 9 is input to the moving picture decoding device 97 via the input buffer 92 and is decoded into a video signal, and then displayed on the video signal display device 94.

As described above, the moving picture coding / decoding apparatus according to the fourth embodiment of the present invention uses the moving picture coding apparatus according to the first embodiment of the present invention to obtain an I-frame. Of the image quality of the GOP and the accompanying loss of the uniformity of the image quality of the GOP are prevented.

Here, the moving picture coding apparatus according to the first embodiment of the present invention is used as the moving picture coding apparatus, but the coding apparatus according to the second embodiment of the present invention is used. Has the same effect.

[0071]

As described above, the present invention provides a GOP remaining amount updating unit, a frame bit generation amount counting unit, a frame target calculating unit, an encoding unit, a GOP remaining amount measuring unit, an offset calculating unit. By providing the virtual buffer occupancy calculating means, it is possible to obtain an advantageous effect that an excellent moving picture encoding apparatus and a superior decoding system capable of preventing the deterioration of the image quality due to the restriction of the buffer underflow can be realized.

[Brief description of the drawings]

FIG. 1 is a block diagram of a frame target calculation unit of a video encoding device according to a first embodiment of the present invention.

FIG. 2 is a block diagram of the video encoding device according to the first embodiment of the present invention.

FIG. 3 is a transition diagram of parameters of a frame target calculation unit of the video encoding device according to the first embodiment of the present invention.

FIG. 4 is a flowchart of a processing procedure of a frame target calculation unit of the video encoding device according to the first embodiment of the present invention.

FIG. 5 is a block diagram of a frame target calculation unit of the video encoding device according to the second embodiment of the present invention.

FIG. 6 is a flowchart of a processing procedure of a frame target calculation unit of the video encoding device according to the second embodiment of the present invention.

FIG. 7 is a block diagram of a video decoding system according to a third embodiment of the present invention.

FIG. 8 is a block diagram of a video decoding system according to a fourth embodiment of the present invention.

FIG. 9 is a block diagram of a video decoding system according to a fourth embodiment of the present invention.

FIG. 10 is a block diagram of a conventional video encoding device.

FIG. 11 is a block diagram of a frame target calculation unit of a conventional video encoding device.

FIG. 12 is a transition diagram of parameters of a frame target calculation unit of a conventional video encoding device.

FIG. 13 is a flowchart of a processing procedure of a frame target calculation unit of a conventional video encoding device.

FIG. 14 is a block diagram of a conventional video decoding system.

FIG. 15 is a diagram illustrating a change in the capacity of a virtual buffer.

[Explanation of symbols]

 1, 13 frame type input terminal 2 video signal input terminal 3 encoding means 4 bit stream output terminal 5 macroblock bit generation amount counting means 6 frame bit generation amount counting means 7 quantization coefficient adding means 8 frame target calculation unit 9 quantization Coefficient calculating means 14 Complexity calculating means 15 Switching means 16-18 Complexity storing means 19 GOP remaining amount calculating means 20 GOP remaining amount storing means 21 Frame target calculating means 22 Virtual buffer occupying amount calculating means 23 Virtual buffer occupying amount storing Means 24 Underflow detection means 25, 55 Overflow detection means 41, 41A Frame target calculation unit 51 Offset calculation means 52 Offset storage means 53 Subtraction means 72, 82 Video encoding device 73 Transmission buffer 74 Transmission Road 75 receive buffers 76,93 moving picture decoding apparatus 83 output buffers 84,91 recording means 92 the input buffer

Claims (13)

[Claims] An I frame, a P frame, and a B frame are repeated in a predetermined order in a GOP unit so that a bit generation amount of a GOP target is obtained according to one of the input frame types. Encoding means for encoding; frame bit generation amount counting means for counting the frame bit generation amount of the moving image encoded by the encoding means; frame bit generation amount counted by the frame bit generation amount counting means From a GOP target to update a GOP remaining amount from a GOP target, and output a bit generation target value of the frame according to the input frame type; and a frame target output from the frame target arithmetic unit. Quantized coefficient calculating means for generating a quantized coefficient for encoding by the encoding means A video encoding apparatus comprising: a GOP remaining amount storage unit configured to store a GOP remaining amount updated based on a frame bit generation amount for each frame from a target GOP target; Offset calculating means for calculating an offset which is a difference between the virtual buffer occupancy at the beginning of each GOP and the virtual buffer occupancy target value; offset storing means for storing the offset calculated by the offset calculating means; Subtraction means for correcting the remaining GOP by subtracting the offset from the remaining GOP stored in the remaining GOP storage means; and setting the target of the bit generation amount in the frame based on the remaining GOP corrected by the subtraction means. And a frame target calculating means for calculating. 2. The video coding apparatus according to claim 1, wherein said video coding apparatus inputs a quantization coefficient from said coding means, and after coding one frame, outputs a sum of said quantization coefficient in one frame. And an encoding unit. The encoding unit has a quantization unit. The frame target calculation unit has a GOP remaining amount stored in the GOP remaining amount storage unit and the offset. 2. The moving picture coding apparatus according to claim 1, wherein a frame target is calculated from an offset stored in a storage means and a sum of quantization coefficients of one frame of said quantization coefficient addition means. 3. The video encoding device detects an underflow of a virtual buffer from a frame target output from the frame target calculation means and a virtual buffer occupancy, and detects the underflow of the frame target when the underflow is detected. 3. The moving picture coding apparatus according to claim 2, further comprising an underflow detecting means for subtracting and outputting to said coding means. 4. The video encoding apparatus detects an overflow of a virtual buffer from a frame target output from the underflow detection means and a virtual buffer occupancy, and increases the frame target when an overflow is detected. 4. The moving picture coding apparatus according to claim 3, further comprising overflow detecting means for outputting to said coding means. 5. The moving picture coding apparatus according to claim 1, further comprising an overflow detecting means for correcting the virtual buffer occupancy to an upper limit value of the virtual buffer when detecting an overflow of the virtual buffer from the virtual buffer occupancy. The moving picture coding apparatus according to claim 3, wherein: 6. The video encoding device according to claim 2, an output buffer for receiving a bit stream output from the video encoding device, and transmission for transmitting the bit stream input to the output buffer. A reception buffer that inputs the bit stream transmitted on the transmission path, and a video decoding device that decodes the bit stream input to the reception buffer and outputs a video signal. A moving picture decoding system characterized by the following. 7. The video encoding device according to claim 3, an output buffer for inputting a bit stream output from the video encoding device, and transmission for transmitting the bit stream input to the output buffer. A reception buffer that inputs the bit stream transmitted on the transmission path, and a video decoding device that decodes the bit stream input to the reception buffer and outputs a video signal. A moving picture decoding system characterized by the following. 8. The video encoding device according to claim 4, an output buffer for receiving a bit stream output from the video encoding device, and a transmission for transmitting the bit stream input to the output buffer. A reception buffer that inputs the bit stream transmitted on the transmission path, and a video decoding device that decodes the bit stream input to the reception buffer and outputs a video signal. A moving picture decoding system characterized by the following. 9. The video encoding device according to claim 5, an output buffer for inputting a bit stream output from the video encoding device, and transmission for transmitting the bit stream input to the output buffer. A reception buffer that inputs the bit stream transmitted on the transmission path, and a video decoding device that decodes the bit stream input to the reception buffer and outputs a video signal. A moving picture decoding system characterized by the following. 10. A moving picture coding apparatus according to claim 2, recording means for recording a bit stream output from said moving picture coding apparatus, and video obtained by decoding said bit stream recorded by said recording means. A video decoding system, comprising: a video decoding device that outputs a signal. 11. A moving picture coding apparatus according to claim 3, recording means for recording a bit stream output from said moving picture coding apparatus, and video obtained by decoding said bit stream recorded by said recording means. A video decoding system, comprising: a video decoding device that outputs a signal. 12. A moving picture coding apparatus according to claim 4, recording means for recording a bit stream output from said moving picture coding apparatus, and a video obtained by decoding said bit stream recorded by said recording means. A video decoding system, comprising: a video decoding device that outputs a signal. 13. The moving picture encoding apparatus according to claim 5, recording means for recording a bit stream output from said moving picture encoding apparatus, and video obtained by decoding said bit stream recorded by said recording means. A video decoding system, comprising: a video decoding device that outputs a signal.