JP2001028753A - Dynamic image coder and its method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a method by which a code amount is assigned more properly between picture types even when an object average bit rate is closer to a maximum transfer rate in a variable bit rate control method adopting a 1-path or 2-path method for a dynamic image coder. SOLUTION: The dynamic image coder encoding a dynamic image is provided with a means 23 that detects a code quantity generated from the dynamic image, a means 22 that detects a mean quantization scale of the dynamic image, a means 25 that detects a coding image characteristic of at least the dynamic image in the dynamic image and a motion compensation prediction image generated by a motion compensation prediction means, a 1st code quantity control means 51 that calculates a 1st assignment code quantity of the image to be coded next on the basis of the generated code quantity and the average quantization scale, a 2nd code quantity control means 52 that calculates a 2nd assigned code quantity placing a limit onto the 1st assigned code quantity and a means 51 that decides the quantization scale of the image to be coded next on the basis of the 1st and 2nd assigned code quantities.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to high-efficiency coding of moving images, and more particularly to a code amount control apparatus and method suitable for performing variable bit rate coding.

[0002]

2. Description of the Related Art MPEG2 has already been defined as an international standard for technology for efficiently encoding moving images such as TV signals. MPEG2 divides a “frame” image that constitutes a moving image into 16 × 16 pixel blocks called “macroblocks”, and for each macroblock unit, a reference image separated by a predetermined number of frames before or after a time frame. A motion amount called a “motion vector” is obtained between encoded images, and “motion compensation prediction” is used to construct an encoded image from a reference image based on the motion amount.
Two types of image coding: a technology and a "transform coding" technology that compresses the amount of information using DCT (Discrete Cosine Transform), which is a type of orthogonal transform, for the error signal of motion compensation prediction or the coded image itself. It is stipulated based on the elemental technology.

FIG. 7 shows an example of a configuration of a conventional MPEG2 moving image encoding apparatus, and FIG. 6 shows an example of an encoded picture structure. In the motion compensated prediction, as shown in the coded picture structure shown in FIG. 7, prediction called I picture (intra-frame coding), P picture (forward prediction coding), and B picture (bidirectional prediction coding) It is composed of a combination of three types of pictures different in method. As shown in FIG. 7, in the transform coding, the output of the subtracter 71 which is the error signal of the motion compensation prediction by the motion compensation predictor 77 is applied to the coded image itself in the case of the I picture and to the coded image itself in the case of the P and B pictures. ,
DCT is performed by the DCT unit 72.

[0004] After the DCT coefficient obtained by the DCT unit 72 is quantized by the quantizer 73 under the control of the output of the code amount control unit 90, it is changed together with other accompanying information such as a motion vector. The long coding is performed by the variable length coder 75, and the coded sequence is output after being stored in the buffer 76 as a “bit stream”. At this time, the quantization scale is controlled by the code amount control unit 90 in accordance with the sufficiency of the buffer 76. On the other hand, the output coefficients of the quantizer 73 are supplied to an inverse quantizer 77 and an IDCT unit 78, locally decoded, and stored in a frame memory 81 for each block.

[0005] Since MPEG2 is variable-length coding, the amount of generated code (bit rate) per unit time is not constant. Therefore, it is possible to control the bit rate to a required bit rate by appropriately changing the quantization scale at the time of quantization in the quantizer 73 for each macroblock. MPEG
2 In Test Model 5, a fixed bit rate control method that makes the generated code amount constant in GOP units has been proposed.

In Test Model 5, a different code amount is assigned depending on the picture type. While assigning the largest amount of code to I-pictures in which intra-frame encoding is performed, the quantization scale of B-pictures in which decoded images are not used for prediction again is set to 1.4 times that of I- and P-pictures. By reducing the amount of code to be allocated,
By reducing the amount of code for the B picture and allocating that much to the I and P pictures for which the decoded image is used for prediction,
Optimization of code amount allocation according to the picture type is performed so that the image quality of the decoded image is constant among the picture types.

[0007] The fixed bit rate control method in Test Model 5 is an effective method for applications requiring a constant transfer rate. However, since almost the same code amount is assigned to any part of the moving image sequence, a sufficient code amount is not provided to a complex scene including a large amount of information, resulting in image quality deterioration. On the other hand, in the case of a simple scene with a small amount of information, the code amount becomes excessive and wasteful, and an appropriate rate control method is used for applications where a variable transfer rate is possible such as DVD-Video. I couldn't say.

[0008] A variable bit rate control method is a rate control method that solves the above problems. JP-A-6-14
No. 1298 discloses an encoding device using variable bit rate control. In this device, first, provisional encoding is performed on an input moving image using a fixed quantization scale, and the generated code amount is counted for each unit time. Next, the target transfer rate of each part is set based on the generated code amount at the time of provisional encoding so that the generated code amount of the entire input moving image becomes a required value. Then, while performing control so as to match the target transfer rate, the second encoding, that is, actual encoding, is performed on the input moving image.

However, in the above conventional example, encoding must be performed at least twice in order to obtain an output bit stream, and in an application requiring real-time performance, a two-pass system such as this apparatus is used. Variable bit rate control cannot be used.

On the other hand, there is also a variable bit rate control method for encoding a moving picture in almost real time, that is, a variable bit rate control method of a one-pass system. Japanese Patent Application Laid-Open No. H10-164577 discloses an encoding apparatus using a one-pass variable bit rate control method, for example, in FIG.

FIG. 8 shows an example of the configuration of the conventional moving picture coding apparatus. The same components as those in FIG. 7 are denoted by the same reference numerals, and description thereof will be omitted. In this conventional device, the code amount stored in the buffer 76 is supplied to the generated code amount detector 83, and the generated code amount by the generated code amount detector 83 and the quantization scale from the quantizer 73 are averaged. Is supplied to the quantization scale detector 82, and the product of the average quantization scale detector 82 and the average value of the quantization scales in the screen is defined as "screen complexity".
4 to determine a target code amount or a target quantization scale of the screen based on the ratio of the current screen complexity to the average value of the past screen complexity, thereby performing variable bit rate control on the code amount controller. 74.

[0012]

However, variable bit rate control is often limited by the maximum transfer rate. If the target average bit rate is sufficiently smaller than the maximum transfer rate, optimize the code amount allocation between picture types by making the code amount allocation for B pictures smaller than I and P pictures as in Test Model 5. Is possible.

When the target average bit rate approaches the maximum transfer rate, the allocated code amounts of the I and P pictures are restricted by the maximum transfer rate, and the difference between the allocated code amounts with the B pictures is reduced. The code amount becomes almost the same between picture types. If the difference in the allocated code amount becomes small, the image quality of the I and P pictures becomes relatively worse than that of the B picture, and this is caused by inappropriate code amount distribution despite the high target average bit rate. There is a problem that image quality degradation is perceived due to the difference in image quality.

Accordingly, the present invention provides a method for controlling a variable bit rate of a one-pass or two-pass scheme in a moving picture coding apparatus, wherein even if a target average bit rate is close to a maximum transfer rate, more appropriate codes can be used between picture types. It is an object to provide a method for realizing a quota.

[0015]

According to the present invention, there is provided a moving picture coding apparatus having motion compensation prediction such as MPEG2, orthogonal transform, quantization, and variable length coding. Means for detecting an average quantization scale, first code amount control means for determining an assigned code amount of an image to be encoded next from the obtained generated code amount and the average quantization scale,
The present invention is characterized in that it has a second code amount control means for limiting the allocated code amount obtained by the first code amount control means for each picture type. For example, when code amount allocation is performed by variable bit rate control, the first code amount control means obtains the allocated code amount by variable bit rate control, and the second code amount control means performs fixed bit rate control of the maximum transfer rate. The upper limit of the allocated code amount is obtained. The code amount actually allocated to each image is set to the upper limit only when the allocated code amount obtained by the first code amount control means exceeds the upper limit of the allocated code amount obtained by the second code amount control means. In other cases, the allocated code amount obtained by the first code amount control means is set as the actual allocated code amount. As a result, even when the target average bit rate is close to the maximum transfer rate, it is possible to optimally hold the code amount allocation between picture types without giving an unnecessarily large code amount to the B picture. .

In the above moving picture coding apparatus, a variable bit rate of a one-pass system is obtained by using a screen complexity obtained by performing a predetermined operation on a product of a generated code amount of each picture and an average quantization scale. When performing control, a predetermined function having the screen complexity as a factor is set, and this function is multiplied by the upper limit of the allocated code amount obtained by the second code amount control means to change the upper limit of the allocated code amount. Accordingly, it is possible to suppress an increase in the code amount in a portion near the upper limit of the allocated code amount, and to alleviate a change in image quality at a point where the upper limit of the allocated code amount is exceeded.

Further, when one coded bit stream is shared by two decoding systems by applying the above moving picture coding apparatus, for example, I, P,
In normal coding using all B pictures, one decoding system decodes the entire bit stream, and the other decoding system decodes only I and P pictures. And the second code amount control means controls the code amount of the bit stream portion of only the I and P pictures, and the code amounts of the I and P pictures can be obtained by the second code amount control means. By performing the code amount allocation of the first code amount control means while limiting to the allocated code amount, it is possible to perform coding that can be decoded by two decoding systems with one coding device.

[0018]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS First Embodiment
Will be described below with reference to FIG. As shown in FIG. 1, a first embodiment of the moving picture coding apparatus and method according to the present invention includes a subtractor 11, a DCT unit 12, a quantizer 13, a variable length coder 15, a buffer 16, Inverse quantizer 17, IDCT unit 18, motion compensation predictor 19, adder 20, frame memory 21, average quantizer scale detector 22, generated code amount detector 23, screen complexity calculator 24,
Image characteristic detector 25, VBR code amount controller 51, and C
A BR code amount controller 52 is provided.

It is assumed that the original image is divided in advance into macroblock units by an image block divider (not shown). With respect to the divided source image, motion compensation prediction is not performed for an I picture, and the source image block itself passes through a subtractor 11 to a DC which is a type of an orthogonal transformer.
After being sent to the T unit 12 and subjected to DCT, it is quantized by the quantizer 13 by the quantization scale sent from the code amount controller 14.

The quantized signal is converted into a code by the variable-length encoder 15 and adjusted by the next buffer 16 to output the code. On the other hand, the output coefficient of the quantizer 13 is locally decoded by the inverse quantizer 17 and the IDCT unit 18, and the output of the motion compensation predictor 19 is not added by the adder 20. Stored in

For the P and B pictures, the divided original moving picture and a predetermined locally decoded picture block stored in the frame memory 21 are supplied to the motion compensation predictor 19, where the motion vector detection and motion compensation are performed. Then, a difference between pixels of the predicted image block and the original image block is calculated by the subtractor 11, and the error image block, which is the difference value, is calculated by DC.
It is sent to the T unit 12.

Thereafter, the DCT is performed in the same manner as the I picture.
The difference value is subjected to DCT by the encoder 12, quantized by the quantizer 13 by the quantization scale sent from the code amount controller 14, converted into a code by the variable length encoder 15, and then converted by the next buffer 16. The code is output after the adjustment.

The output coefficient of the quantizer 13 is the inverse quantizer 1
After local decoding by the IDCT unit 7 and the IDCT unit 18, the predicted image block from the motion compensation predictor 19 is added for each pixel by an adder 20, and is stored in the frame memory 21 for each block. For each picture, the quantizer 13 sends the quantization scale for each macroblock to the average quantization scale detector 22, where the quantization scale for one frame is added, and the average quantization scale for one frame is calculated. Is calculated.

On the other hand, the generated code amount is monitored in the buffer 16, and the value is sent to the generated code amount detector 23. The generated code amount detector 23 adds the generated code amount for each frame, and detects the generated code amount for one frame. The average quantization scale and the generated code amount detected for each frame are sent to the screen complexity calculator 24 and the CBR code amount controller 52, respectively.

On the other hand, the image characteristic detector 25 receives the divided original image, detects a parameter indicating an image characteristic for each frame of the original image in units of macroblocks, that is, an activity, adds the parameters for each frame, and adds the activity. The result is sent to the screen complexity calculator 24.

That is, as for the input to the image characteristic detector 25, since motion compensation prediction is not performed in the case of an I picture, only the original moving image divided in macroblock units is input, and the image characteristics are determined in macroblock units. The activity (ACTcur), which is a parameter to be indicated, is detected, added for each frame, and sent to the screen complexity calculator 24 as an I-picture activity ACTi.

As the activity (ACTcur), variance of luminance value, difference value between pixels and the like can be considered, but other parameters may be used as long as they show image characteristics.

On the other hand, when the input to the image characteristic detector 25 shown in FIG. 1 is a P or B picture, in addition to the divided original image, an error image or a motion vector in the motion compensation prediction in macroblock units is detected. , And a motion vector used in the motion compensation prediction from the motion compensation prediction unit 19. As in the case of the I picture, the (original image) activity ACTcur is detected for each macroblock from the divided original image.

On the other hand, the error image in the motion compensation prediction in units of macroblocks or the difference image between the coded image in the motion vector detection and the reference image is subjected to the sum of absolute values or the sum of squared errors in the error image.
Is detected as Further, the motion vector used in the motion compensated prediction has an absolute value of a difference between each component and an adjacent macroblock, and is detected as ACTmv.

Then, ACTmb is calculated for each macroblock by the operation of the following equation (1), and ACTmb is added for one frame, and the activities ACTp and AC of the P and B pictures are calculated.
It is sent to the screen complexity calculator 24 as Tb.

ACTmb = a · ACTcur + b · ACTpred + c · ACTmv (1)

The values of the constants a, b, and c are changed for each picture, each macroblock prediction mode (intra, unidirectional prediction, or bidirectional prediction). For example, in the case of intra, prediction is not performed like an I picture.
Since b = c = 0, and it is considered that the generated code amount is larger than that of the block to be predicted, the value of a is increased.

As described above, by detecting the activity in accordance with the prediction mode or the like, it is possible to estimate the screen complexity more in accordance with the encoding characteristics.

The screen complexity calculator 24 multiplies the supplied average quantization scale of each frame by the generated code amount, performs a predetermined conversion on the multiplication result, and performs
Required as (past) screen complexity. The screen complexity is calculated by adding the value within a certain period for each coded picture type and then dividing by the number of frames of the same picture type within that period, and the average screen complexity of each picture type Xi-ave, X
p-ave and Xb-ave are calculated.

During the certain period, the number of frames predetermined in time before the picture for which coding has just been completed, for example, a certain number of frames such as 15 frames or 300 frames, may be used. In some cases, such as from a frame to an image for which encoding has just finished, the number of frames is sequentially increased. Note that, even in the case of the former fixed number of frames, if the number of encoded frames does not satisfy the predetermined certain period, the number of frames is sequentially increased similarly to the latter.

The screen complexity Xk-c (k = i, p, b) of the current image to be encoded is obtained by encoding the activity of the current image immediately before ACTk (k = i, p, b). The screen complexity Xk-p (k = i, p, b) and the activity ACTk-p (k = i, p, b) of an image of the same picture type can be estimated by the following equation (2).

Xk-c = Xk-p · ACTk / ACTk-p (2)

In the initial state, if there is no frame for which the same picture type has been coded, the screen complexity and activity of each picture type picture are obtained in advance for some pictures, and these are averaged. The average value may be statistically averaged in accordance with the frequency of occurrence of a moving image, and may be used as an initial value.

Average screen complexity of each picture type Xi-av
e, Xp-ave, Xb-ave and the estimated screen complexity Xi-c, Xp-c, Xb-c of the current image to be encoded are sent to the VBR code amount controller 51, where the variable bit rate The quantization scale for control is set. Target average bit rate is BitRate, number of frames per second is PictureRate, 1
Assuming that the number of frames in one GOP (usually the interval between I pictures), which is one coding unit, is N, the average allocated code amount Rave of one GOP is given by the following equation (3).

Rave = (BitRate / PictureRate) · N (3)

If Rave in the above equation is the required allocated code amount of 1 GOP at the time of the average screen complexity, the 1 GOP image including the current image to be coded from now on is uniformly calculated by the screen complexity calculator 24.
Assuming that it is equal to the estimated screen complexity of the current image obtained in the above, the required allocated code amount Rck (k = i, p, b) of 1 GOP required to keep the image quality constant is given by the following equation (4). Given.

Rck = Rave · Xk-c / Xk-ave (4)

By appropriately allocating Rck (k = i, p, b) in the above equation to each picture of 1 GOP, the target code amount in the first code amount control means of the current image to be encoded is calculated. . As an example, a target code amount allocation method of MPEG2 Test Model 5 is described below, but other methods may be used.

It is assumed that the number of frames of P and B pictures included in one GOP is Np and Nb, and the quantization scale setting ratio of P and B pictures to I picture is Kp and Kb. At this time, the target allocation code amounts Ti, Tp, and Tb of each picture type are expressed by the following equations (5), (6), and (7).
Given by Note that MAX [a, b] indicates an operation of selecting the larger one of a and b. Xi, Xp, and Xb are the screen complexity (product of the average quantization scale of the picture and the generated code amount) of the picture coded immediately before.

(I picture) Ti = MAX [A, B] A = Rc / (1 + Np.Xp / (Xi.Kp) + Nb.Xb / (Xi.Kb)) B = BitRate / (8.PictureRate ) (Five)

(P picture) Tp = MAX [C, D] C = Rc / (Np + NbKpXb / (KbXp)) D = BitRate / (8PictureRate) (6)

(B picture) Tb = MAX [E, F] E = Rc / (Np + NpKbXp / (KpKb)) F = BitRate / (8PictureRate) (7)

On the other hand, the CBR code amount controller 52 receives the average quantization scale and the generated code amount for each frame, and calculates the screen complexity Xi,
BitRate is the highest transfer rate (BitRateMax) in search of Xp and Xb
Target code amount Ti-max, Tp-m for each picture type
ax, Tb-max are obtained in the same manner as the target assigned code amounts Ti, Tp, Tb in the first code amount control means. Here, the average allocated code amount Rav-max of one GOP is common to each picture and is given by the following equation (8).

Rav-max = (BitRateMax / PictureRate) · N (8)

(I picture) Ti-max = MAX [A, B] A = Rav-max / (1 + Np.Xp / (Xi.Kp) + Nb.Xb / (Xi.Kb)) B = BitRateMax / (8 · PictureRate) (9)

(P picture) Tp-max = MAX [C, D] C = Rav-max / (Np + Nb · Kp · Xb / (Kb · Xp)) D = BitRateMax / (8 · PictureRate) (10)

(B picture) Tb-max = MAX [E, F] E = Rav-max / (Np + NpKbXp / (KpKb)) F = BitRateMax / (8PictureRate) (11)

In the above equation, Ti-max, Tp-max, and Tb-max are the upper limits of the target allocated code amount in the second code amount control means, and these values are transmitted to the VBR code amount controller 51. The value of Ti, Tp, Tb is Ti-max, Tp-ma for the picture type of the current image to be encoded.
A limiter is applied based on the values of x and Tb-max, and the target assigned code amount of the current image is determined.

Based on the target allocated code amount determined as described above and the generated code amount of each macroblock detected in the buffer 16, the quantization scale of each macroblock is determined using the MPEG2 Test Model 5 method. To determine.

The activity ACTcur of each macroblock is also transmitted from the image characteristic detector 25 to the code amount controller 51.
Is transmitted and used for adaptive quantization control that changes the quantization scale of each macroblock based on the activity in MPEG2 Test Model 5, but this adaptive quantization control may not be performed. Alternatively, the quantization scale of each macroblock may be determined by a completely different method.

The quantization scale of each macroblock output from the code amount controller 51 is sent to the quantizer 13, and the current image (an original image after DCT or an error image block for motion compensation prediction) is converted After being quantized by this quantization scale, variable-length coded and adjusted by the buffer 16, a code is output.

The quantization scale for each macroblock and the generated code amount monitored by the buffer 16 are sent to the average quantization scale detector 22 and the generated code amount detector 23, respectively, and are used for controlling the code amount of the next picture. Is done.

In the above description, the CBR code amount controller 52
, The average allocated code amount Rav-max of one GOP is simply calculated as the code amount allocated to one GOP at the maximum transfer rate (BitRateMax). On the other hand, a predetermined function as shown in FIG. 2 (a), which is based on the estimated screen complexity Xk-c (k = i, p, b) of the current image to be encoded, When the estimated screen complexity Xk-c of the image increases, a function f (Xk-c) whose value approaches 1 infinitely is set.

By using Rav-max ′ in the following equation (12) obtained by multiplying this function for each picture type instead of Rav-max, as shown in FIG. Along with gradually reducing the amount of generated code when the transfer rate is close to the transfer rate, image quality degradation of the picture type caused by the discontinuity of the relationship between the screen complexity and the allocated code amount at the point where the maximum transfer rate is exceeded is remarkable. It is also possible to suppress certain problems.

Rav-max ′ = f (Xk−c) · (BitRateMax / PictureRate) · N where f (Xk−c) is a function (k = i, p, b) shown in FIG. )

Next, the second embodiment of the moving picture coding apparatus of the present invention will be described.
Will be described below with reference to FIG. In the second embodiment, the present invention is applied to variable bit rate code amount control of the two-pass system. The basic encoding portion from the input of the original image to the conversion to the code by the variable length encoder 15 is the same as that of the first embodiment. The major difference is that, for one image, the encoding operation is performed twice (or more), the temporary encoding is performed first, and the second encoding is performed based on the result of the generated code amount. And that there is a difference between the two encoding operations.

In the first encoding, the quantization scale sent to the quantizer 13 is not sent from the VBR code amount controller 51 but is fixed from the provisional encoding quantization scale setting unit 56 via the switch SW1. (A value such as 6 or 8) is sent, and thereby a fixed value is quantized. Then, the bit stream after the variable length encoding is performed in the variable length encoder 15 is not sent to the buffer 16 for outputting the bit stream to the outside, and the provisional encoding generated code amount detector 53 is output via the switch SW2. And the generated code amount of each image in the first encoding is detected.

The generated code amount is sequentially sent from the temporary encoding generated code amount detector 53 to the temporary transfer rate memory 54, and is added every predetermined period to calculate the temporary transfer rate. This operation is performed until the encoding of one image sequence is completed, and the temporary transfer rate memory 54 stores the temporary transfer rate for each predetermined period.

When the first encoding is completed, the temporary code amount or the average temporary transfer rate of the entire image sequence is calculated, and this value and the temporary transfer rate for each predetermined period are sent to the target transfer rate calculator 55. A target transfer rate for each predetermined period in the second encoding (actual encoding) is calculated.

The provisional transfer rate Rt for each predetermined period in the first encoding and the target transfer rate R for the second encoding
Is set in advance with a predetermined function. For example, the following function (13) can be considered.

R = a · (Rt) ^b (where a and b are constants, a> 0, 0 <b <1) (13)

When the first provisional encoding is completed and the target transfer rate for the second encoding is determined, the second encoding (actual encoding) is started according to the target transfer rate. In the second encoding, the value obtained by the VBR code amount controller 51 is sent to the quantization scale sent to the quantizer 13.

Here, in the VBR code amount controller 51, the average quantization scale and the generated code amount of each frame detected by the average quantization scale detector 22 and the generated code amount detector 23, and the target transfer rate calculator From 55, the target assigned code amount of the image to be encoded is obtained from the target transfer rate for each predetermined period calculated from the provisional encoding result.

On the other hand, the CBR code amount controller 52 calculates
The upper limit of the target allocated code amount (in the second code amount control means) is also input to the VBR code amount controller 51, and the target allocated code amount is limited to determine the target allocated code amount.
The upper limit of the target allocation code amount in the CBR code amount controller 52 is Ti-max, Tp-ma (based on Rav-max) in the first embodiment.
Same as x, Tb-max.

Based on the target allocated code amount determined as described above and the generated code amount of each macroblock detected in the buffer 16, the MPEG2 Test Mod as in the first embodiment.
The quantization scale of each macroblock is determined using a method such as el5.

The quantization scale of each macroblock determined in this way is sent to the quantizer 13, and an image to be coded therefrom (an original image after DCT or an error image block for motion compensation prediction) is obtained. Quantization is performed using this quantization scale and variable-length encoding is performed.

The bit stream generated here is supplied to the buffer 16 in the second encoding, where the bit stream is adjusted by the target transfer rate for each predetermined period calculated by the target transfer rate calculator 55, and then the code is output. Is done. The quantization scale for each macroblock and the generated code amount monitored by the buffer 16 are sent to the average quantization scale detector 22 and the generated code amount detector 23, respectively, and used for the code amount control of the next picture.

Further, a third embodiment of the moving picture coding apparatus according to the present invention will be described below with reference to FIGS. The first and second embodiments described above are cases where the present invention is applied to variable bit rate code amount control, but the present invention is not limited to this, and can be applied to a wide range of applications. As in the moving picture coding apparatus shown in FIG. 4, the coded bit stream output from the buffer 16 is divided into two by a stream divider 59. One output outputs the entire coded bit stream using all the I, P, and B pictures as shown in FIG. 5, and the other outputs the coded bit stream using only the I and P pictures.

It is assumed that the output of only the I and P pictures is a header converter (not shown) in which the parameters and the like of the header portion have been rewritten to appropriate values. The code amount controller 1 (51A) of FIG. 4 controls the code amount of the entire bit stream, and the code amount controller 2 (52A) controls the code amount of the bit stream of only I and P pictures.

Here, the code amount controller 1 (51A) and the code amount controller 2 (52A) set the average bit rates of both picture types so that the average allocated code amount of each picture type is relatively close. It is assumed that In order to simultaneously satisfy the code amount control of the two bit streams, the code amount control result of the I and P pictures in the code amount controller 2 (52A) is sent to the code amount controller 1 (51A).
In (51A), the code amount allocation of the I and P pictures resulting from the code amount controller 2 (52A) is directly applied to the I and P pictures, and the code amount is newly allocated to the B picture and the code amount control is performed. Control of the container 1 (51A) is realized.

As a result, the entire bit stream is controlled by the code amount controller 1 (51A).
When the I / P picture portion of the bit stream is extracted by the stream divider 59, the code amount controller 2 (52A)
The bit stream controlled by can be obtained.

When it is not necessary for the stream divider 59 to output a bit stream consisting of only I and P pictures, the code amount controller 1 (52A) does not perform code amount allocation, and the code amount controller 1 ( 51A), I, P, B
Normal code amount allocation is performed for each picture. When the code amount is not allocated by the code amount controller 2 (52A), a signal indicating this is sent to the stream divider 59, and the output of the bit stream of only I and P pictures is stopped.

The present invention is not limited to the embodiment shown in FIG. 4, but in the code amount control mode having two code amount controllers, the main code amount controller 51A is replaced with the subordinate code amount controller 52A. Of the picture amount of each picture type by applying a restriction based on the result of the code amount assignment for each picture type, or a moving picture corresponding to two decoding systems for one encoded bit stream. An encoding device can be realized.

[0079]

As described above, according to the present invention, the first code amount control means for determining the allocated code amount of the coded image, and the second code amount control means for limiting the code amount allocation of the first code amount control means. For example, in the case of variable bit rate control, the first code amount control means uses a variable bit rate control, and the second code amount control means uses a fixed bit rate control with a maximum transfer rate. And the actual allocated code amount is determined by the second code amount control means only when the allocated code amount obtained by the first code amount control means exceeds the allocated code amount obtained by the second code amount control means. Is applied, otherwise, the assigned code amount obtained by the first code amount control means is applied. Thus, even when the target average bit rate is close to the maximum transfer rate,
It is possible to optimally maintain the code amount allocation between picture types without giving an unnecessarily large code amount to the B picture.

When the control using the screen complexity obtained by performing a predetermined operation on the product of the generated code amount of each image and the average quantization scale is performed by the variable bit rate control of the 1-pass method, Set a predetermined function with degree as a factor,
This function is multiplied by the upper limit of the allocated code amount obtained by the second code amount control means to change the upper limit of the allocated code amount, thereby suppressing an increase in the code amount in a portion near the upper limit of the allocated code amount. At the same time, it is possible to mitigate image quality fluctuation at a point where the allocated code amount exceeds the upper limit.

Also, in the case where one coded bit stream is shared by two decoding systems by applying the above-mentioned moving picture coding apparatus, the coding amount allocation obtained by the second coding amount control means can reduce By limiting the code amount allocation of one code amount control unit, it becomes possible to perform coding that can be decoded by two decoding systems with one coding device.

[Brief description of the drawings]

FIG. 1 is a first embodiment of a moving picture coding apparatus and method according to the present invention.
FIG. 2 is a block diagram showing an embodiment of the present invention.

FIG. 2 is a diagram illustrating a function for calculating Rav-max ′ and a relationship between Xk-c and an allocated code amount according to the first embodiment of the present invention.

FIG. 3 shows a second example of the moving picture coding apparatus and method according to the present invention.
FIG. 2 is a block diagram showing an embodiment of the present invention.

FIG. 4 is a third embodiment of the moving picture coding apparatus and method according to the present invention;
FIG. 2 is a block diagram showing an embodiment of the present invention.

FIG. 5 is a diagram illustrating a state of bit stream division by a stream division unit according to a third embodiment of the present invention.

FIG. 6 is a diagram illustrating an example of a coded picture structure.

FIG. 7 is a diagram illustrating a configuration example of a general moving image encoding device.

FIG. 8 is a diagram illustrating an example of a configuration of a conventional video encoding device.

[Explanation of symbols]

Reference Signs List 11 subtracter 12 DCT unit (orthogonal transformer) 13 quantizer 14 code amount controller 15 variable length encoder 16 buffer 17 inverse quantizer 18 IDCT unit 19 motion compensation predictor 20 adder 21 frame memory 22 average quantum Scaled scale detector 23 generated code amount detector 24 screen complexity calculator 25 image characteristic detector 51 VBR code amount controller 52 CBR code amount controller 51A code amount controller 1 52A code amount controller 2 53 provisional coding generation Code amount detector 54 Temporary transfer rate memory 55 Target transfer rate calculator 56 Temporary coding quantization scale setting unit 59 Stream splitter Rav-max Average allocated code amount of 1GOP when BitRate is at maximum transfer rate (BitRateMax) Rck 1GOP Required allocated code amount (k = i, p, b) SW1, SW2 switch Tk Target allocated code amount of each picture type (k = i, p, b) Tk-max BitRate is the highest transfer rate (BitRateMax), target allocated code amount of each picture type (k = i, p, b) Xk Screen complexity of current image (k = i, p, b) Xk-c Estimated screen of current image Complexity (k = i, p, b)

 ──────────────────────────────────────────────────続 き Continued on the front page F term (reference) 5C059 KK01 KK22 MA00 MA23 MC11 MC38 ME01 NN01 NN28 PP05 PP06 PP07 SS11 TA60 TC02 TC10 TC18 UA02 UA33

Claims (9)

[Claims] 1. A moving picture coding apparatus for coding an input moving picture by a motion compensation predicting means, an orthogonal transform means, a quantizing means, and a variable length coding means, comprising the steps of: Means for detecting a code amount; means for detecting an average quantization scale of each image of the input moving image; and at least the input moving image of the input moving image and the motion-compensated predicted image generated by the motion-compensated predicting means. Means for detecting a coded image characteristic of an image; and a generated code amount detected by the means for detecting the generated code amount and an average quantization scale detected by the means for detecting the average quantization scale. A first code amount control unit for calculating a first allocated code amount of an image to be converted, and a second code amount calculating unit for calculating a second allocated code amount for limiting the first allocated code amount. A moving picture, comprising: a code amount control means; and a means for determining a quantization scale of the next image to be coded from the allocated code amounts calculated by the first and second code amount control means. Image coding device. 2. The moving picture coding apparatus according to claim 1, wherein an upper limit of the allocated code amount is calculated for each picture type (I picture, P picture, B picture), When determining the actual allocated code amount in the first code amount control means, the calculated first allocated code amount exceeds the upper limit of the allocated code amount set in the second allocated code amount. If the second
A moving image coding apparatus characterized in that, in other cases, the first allocated code amount is determined as an actual allocated code amount. 3. The moving picture coding apparatus according to claim 1, wherein said second code amount control means is a fixed bit rate code amount control method, and wherein said first code amount control is performed. The means is a variable bit rate code amount control method. 4. The moving picture coding apparatus according to claim 1, wherein a generated code amount, an average quantization scale, and a coded image characteristic of each of the detected images are detected. Means for calculating screen complexity from image characteristic parameters detected by the means, wherein the second assigned code amount is changed by a predetermined function having the screen complexity as a factor. Encoding device. 5. A moving picture coding apparatus according to claim 1, further comprising a stream dividing means for extracting a part of the output code string coded by said variable length coding means. The first code amount control means controls the code amount of the entire output code string, and the second code amount control means determines a code amount of a part of the output code string extracted from the stream division means. A moving picture coding apparatus characterized by controlling. 6. A moving picture coding method for coding an input moving picture by a motion compensation prediction step, an orthogonal transformation step, a quantization step, and a variable length coding step. Detecting a code amount; detecting an average quantization scale of each image of the input moving image; and at least the input moving image of the input moving image and the motion compensated prediction image generated by the motion compensation prediction step. Detecting the encoded image characteristic of the image; and the generated code amount detected by the step of detecting the generated code amount; and the average quantization scale detected by the step of detecting the quantization scale. A first code amount control step of calculating an allocated code amount of an image to be performed, and a step of restricting the first allocated code amount. A second code amount control step of calculating a second allocated code amount; and a quantization scale of the next image to be encoded is determined from the allocated code amounts calculated in the first and second code amount control steps. A moving image encoding method. 7. The moving picture coding method according to claim 6, wherein an upper limit of the allocated code amount is calculated for each picture type (I picture, P picture, B picture), When determining the actual allocated code amount in the first code amount control step, the calculated first allocated code amount exceeds the upper limit of the allocated code amount set by the second allocated code amount. If the second
A moving picture coding method characterized in that, in other cases, the first allocated code amount is determined as an actual allocated code amount. 8. The moving picture coding method according to claim 6, wherein the detected code amount, the average quantization scale, and the coded image characteristics of each of the detected images are detected. Calculating the screen complexity from the image characteristic parameters to be obtained, and changing the second allocated code amount by a predetermined function having the screen complexity as a factor. . 9. The moving picture encoding method according to claim 6, further comprising a stream dividing step of extracting a part of the output code string encoded by said variable length encoding step. The first code amount control step controls the code amount of the entire output code string, and the second code amount control step calculates a code amount of a part of the output code string extracted from the stream division unit. A moving picture coding method characterized by controlling.