JPH06351006A - Variable rate encoding device for picture signal - Google Patents

Abstract

PURPOSE:To flexibly and speedily change an encoding rate with respect to an encoding rate instruction signal from a network side by providing an encoding rate control part which operates control for changing the use rate of each prediction mode in an encoding means, and increasing the rate of the bidirectional prediction encoding. CONSTITUTION:This device is equipped with an encoding rate control part 13 which receives an encoding rate instruction signal (d), and operates control for changing the use rate of each prediction mode in the encoding means, and improving the rate of the bidirectional prediction encoding when the reduction of the encoding rate is instructed by the encoding rate instruction signal (d). Then, at the time of reducing the encoding rate, the encoding rate control part 13 controls the encoding means so that the bidirectional prediction encoding using prediction from preceding and succeeding frames can be used. Therefore, the encoding rate can surely be reduced by suppressing the increase of a quantization noise to the minimum, and using the redundancy of a time direction. Also, the encoding rate can be flexibly changed without operating the switching of a quantization table.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a variable rate coding apparatus for image signals, which is capable of changing the coding rate in a digital transmission apparatus for image signals.

[0002]

2. Description of the Related Art In a system for digitally transmitting an image signal by connecting to a network capable of variable rate transmission,
A plurality of quantization tables having different quantization step sizes are prepared in the encoding device, and the quantization table is switched according to the congestion state of the network to change the encoding rate.

FIG. 3 shows an example of the configuration of a conventional variable rate coding apparatus for image signals. In the figure, quantization table 3
1 is a table including a large quantization step size to a small quantization step size. The quantization control unit 32 selects one quantization table, and the quantizer 33 is set with the coding rate according to the quantization step size. The input image signal a is encoded by an encoding unit including a quantizer 33, a subtractor 34, an adder 35, and a frame memory 36. Buffer memory 37
Temporarily stores the encoded data b and then sends it to the network.

Here, the encoding processing in the encoding unit will be described. Intra-frame encoding is used for encoding processing,
Forward predictive coding using prediction from the previous frame,
There is bidirectional predictive coding that uses prediction from preceding and following frames.

Intra-frame coding is for coding using only the data within the frame, and can be decoded using only the coded data for the frame. The forward predictive coding normally performs motion compensation on the decoded data of the previous frame by shifting the position by the motion vector, and then encodes the difference (prediction error) data with the image sample value at the same position. Is.

In bidirectional predictive coding, normally, after similar motion compensation is performed on decoded data of each of a front frame and a back frame, difference data from an image sample value at the same position in a previous frame or a subsequent frame, or before and after The difference data from the average value of the image sample values at the same position in the frame is encoded.

Each of these encoding processes is performed by the frame memory 3
However, the allocation (use ratio) of the prediction mode to the input frame is fixed in advance. Hereinafter, regarding the allocation of the prediction mode, the MPE shown in FIG.
This will be described with reference to an example of prediction processing by the G (Moving Picture Expert Group) method. The MPEG system has become one of the international standards for moving picture coding in which a plurality of moving picture frames are coded as a set.

In FIG. 4, (1) is bidirectional predictive coding for medium 1 frame, and (2) is bidirectional predictive coding for medium 2 frame. Note that I
Is the intra-frame coded data, and the image signal can be restored only with the I frame data. P is forward predictive coded data, and since predictive coding is performed using I frame data or P frame data, reproduction of these data is necessary for reproduction. B is bidirectional predictive encoded data, which is I frame data and P frame data,
Alternatively, since predictive coding is performed using the preceding and following P frame data, it is necessary to reproduce these data for reproduction. (3) shows an example of the transmission order of each data by the predictive coding of (2), and I, P, B, B, P, ..., I, according to the order of decoding processing on the receiving side. It is transmitted in the order of B and B. Each data includes side information used for prediction of a sync signal, a data type, a motion vector, etc. at a fixed position, detects a delimiter in the middle of the data, extracts the information, and decodes encoded data. Is possible.

In the conventional apparatus, as shown in FIG. 4 (1) or (2), the prediction mode is switched according to the usage ratio preset in the processing frame unit (15 frames), and (3)
As shown in FIG. 7, the transmission is performed in the corresponding order. Here, the coding rate for each of the I, P, and B frames is highest in the I frame data and lowest in the B frame data, as shown in (4). Therefore, if the usage rate of bidirectional predictive coding is increased, the average coding rate can be reduced, but in the past, this usage rate was fixed.

Therefore, the conventional apparatus uses the quantization table 31.
Is changed to change the coding rate. That is, if a quantization table with a large quantization step size is used for the same input signal, the number of output quantization levels decreases, and the code length associated therewith can also be shortened, resulting in a reduction in the coding rate. is there. Conversely, if a quantization table with a small quantization step size is selected, the coding rate will increase.

The quantization table 31 is selected under the control of the quantization controller 32. The quantization controller 32 monitors the data amount c in the buffer memory 37,
When the data amount is larger than a predetermined value, a quantization table with a large quantization step size is selected, and when the data amount is smaller than a predetermined value, a quantization table with a small quantization step size is selected, thereby performing a predetermined encoding. Control the rate.

Further, the quantization controller 32 receives the coding rate instruction signal d from the network and controls the coding rate by using the quantization table of the corresponding quantization step size. For example, when the network becomes congested, the coding rate instruction signal d for reducing the coding rate is received from the network side, and the coding rate is reduced by switching to the quantization table having a large quantization step size. It has become. In addition, the coding rate instruction signal d may include an instruction to increase the coding rate or set a predetermined coding rate in response to the congestion state of the network.

[0013]

As described above, in the conventional variable rate coding apparatus for image signals, the allocation (use ratio) of the prediction mode is fixed, and the coding rate related to the prediction process becomes constant. Was there. Therefore, the coding rate is changed by switching to a quantization table having a different quantization step size.

However, although the coding rate can be reduced by selecting a quantization table having a large quantization step size, deterioration of image quality cannot be avoided because quantization noise increases. That is, in the method of changing the coding rate by changing the quantization step size, it is difficult to significantly reduce the coding rate because the image quality is greatly affected. Further, in the method of transmitting the quantization table itself from the encoding device side to the decoding device side, the information of the quantization table itself must be transmitted along with the change of the quantization table, and the encoded data corresponding to that must be transmitted. Cannot be transmitted. This would rather deteriorate the image quality, and the quantization table could not be changed frequently in order to reduce the coding rate. For this reason, the conventional apparatus cannot flexibly change the coding rate, and its variable range is limited.

Further, the quantization table cannot be changed for the data being encoded. Therefore, a processing delay occurs from the instruction to reduce the encoding rate until the encoding rate is actually reduced, and during that time, data is discarded due to network congestion. Also, the problem of processing delay is the same for encoded data already stored in the buffer memory, and even if the quantization table is switched quickly, the actual encoding rate changes in the buffer memory. It was after all the accumulated data of was transmitted. As described above, in the conventional apparatus, the response to the instruction to change the coding rate is slow, and the delay in the buffer memory is added, so that the data may be discarded on the network during that time to deteriorate the image quality.

SUMMARY OF THE INVENTION It is an object of the present invention to provide a variable rate coding device for an image signal, which can flexibly and quickly change the coding rate in response to a coding rate instruction signal from the network side. .

[0017]

The invention according to claim 1 is provided with an encoding means, a quantization control means and a buffer memory, and the quantization control means has a prediction error by bidirectional predictive encoding with respect to the encoding means. In a variable rate coding device for image signals that allocates a quantization table having a large quantization step size to data, when receiving a coding rate instruction signal and instructing reduction of the coding rate,
It is characterized by further comprising a coding rate control means for changing the usage rate of each prediction mode in the coding means to increase the bidirectional prediction coding rate.

According to a second aspect of the present invention, in a variable rate coding device for an image signal, which is provided with a coding means, a quantization control means and a buffer memory, a coding rate instruction signal is received, which is a coding rate When instructing reduction,
Regarding inter-frame predictive coded data, a coding rate control means for sending only the side information used for the prediction from the buffer memory is provided.

[0019]

According to the first aspect of the present invention, when the coding rate is reduced, the coding rate control means controls the coding means so that the bidirectional predictive coding using the prediction from the preceding and succeeding frames is frequently used. To do. As a result, the coding rate can be reliably reduced by utilizing the redundancy in the time direction while suppressing the increase of the quantization noise to the minimum. Moreover, since the quantization table is not switched, the coding rate can be flexibly changed.

According to a second aspect of the present invention, when the coding rate is reduced, the coding rate control means sends the side information used for the prediction of the inter-frame prediction coded data only to the buffer. Control memory.
This allows the coding rate to be reduced directly,
It is possible to eliminate the delay in changing the coding rate due to the buffer memory. Further, it is possible to promptly respond to the coding rate reduction instruction from the network side.

[0021]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS FIG. 1 shows the configuration of an embodiment of a variable rate coding apparatus for image signals according to the present invention. In the figure, the same parts as those of the conventional one shown in FIG. That is, the quantization table 31, the quantization control unit 32, the quantizer 33, the subtractor 34, and the adder 35 are the same as the conventional one.

The frame memory 11 of this embodiment has a function of changing the usage rate of each prediction mode of intraframe coding, forward prediction coding, and bidirectional prediction coding according to the prediction mode designation signal e. In addition, the buffer memory 12
Has a function of selecting and transmitting only the side information used for the prediction of the inter-frame predictive coded data according to the data selection instruction signal f. The former is a function corresponding to the invention of claim 1, and the latter is a function of claim 2.
Although the functions correspond to the invention described in 1 above and are independent functions, the configuration of the present embodiment is a combination of both functions.

Therefore, the coding rate control unit 13 of the present embodiment receives the coding rate instruction signal d from the network and outputs the prediction mode designation signal e or the data selection instruction signal f for setting the coding rate. The coding rate control unit 13 can also control the coding rate according to the data amount c in the buffer memory 12.

The operation of the apparatus of this embodiment will be described below. When the coding rate control unit 13 receives the coding rate instruction signal d for reducing the coding rate from the network, for example, from the prediction mode for bidirectionally predicting one frame shown in FIG. The prediction mode designation signal e for changing the middle two frames shown in 2) to the prediction mode for bidirectional predictive coding is sent to the frame memory 11. In the frame memory 11, the coding rate can be reduced by increasing the ratio of bidirectional predictive coding (B frame) according to the prediction mode designating signal e. In addition,
At this time, even if the quantization table having the large quantization step size is used, the deterioration of the image quality can be suppressed because the data amount of the B frame is small.

When the coding rate control unit 13 receives the coding rate instruction signal d for reducing the coding rate from the network, it sends the data selection instruction signal f to the buffer memory 12. The buffer memory 12 selects one of the data accumulated according to the data selection instruction signal f.
For inter-frame predictive coded data, by sending only the side information used for the prediction, the coding rate can be quickly reduced.

As described above, according to the present invention, since the bidirectional predictive coding with a small amount of information allocation is heavily used and the coding rate is reduced by reducing the coded data, the response to the temporal change of the image signal may be deteriorated. There is a nature. That is, when the network is congested, the change speed of the image becomes slower, but it can be said that the image quality is less unnatural as compared with the deterioration of the image quality due to the increase of the quantization noise.

As in the conventional case, the quantization control unit 32 selects a quantization table having a quantization step size according to the data amount c in the buffer memory 12 and controls it so that a predetermined coding rate is obtained. It has a function. It also has a function of receiving the coding rate instruction signal d from the network and controlling the coding rate using the quantization table of the corresponding quantization step size. These conventional functions can be used together with the functions according to the present invention described above. Alternatively, the coding rate may be first reduced by the function according to the present invention, and the conventional function may be activated only when the network instructs to reduce the coding rate.

FIG. 2 shows a configuration example of a decoding device corresponding to the encoding device of the present invention. In the figure, the received encoded data b is temporarily stored in the buffer memory 21, and the quantization table and the prediction mode information used for the encoding process are extracted for each frame. The configuration for decoding the encoded data b is the same as that of the encoding device except that the quantizer 33 replaces the inverse quantizer 22. That is, the inverse quantizer 22, the subtractor 34, the adder 35, and the frame memory 1
1 constitutes a decoding unit. The inverse quantizer 22 performs inverse quantization using the quantization table 31 of the quantization step size used in the encoding process. Further, the same prediction mode as in the encoding process is set in the frame memory 11 by the frame control unit 23. With the above configuration, the decoded image signal a is taken out from the output of the adder 35.

[0029]

As described above, according to the present invention, the coding rate can be changed reliably and flexibly while suppressing an increase in quantization noise to a minimum. Furthermore, it is possible to eliminate the delay of the coding rate change due to the switching of the quantization table and the buffer memory, and to promptly respond to the coding rate reduction instruction from the network side. Therefore, it is possible to prevent data discarding due to congestion on the network and stabilize the image quality in the variable rate coding method.

In the conventional method, the I frame data may be discarded on the network, the entire set of frame data may become unreproducible, and the deterioration state of the image quality may be prolonged. On the other hand, in the present invention, the interframe predictive coded data is reduced to only the side information to reduce the coding rate, so that the discarding of I frame data on the network can be prevented and the decoding can be performed. It is possible to minimize the deterioration of image quality in the digitization process.

[Brief description of drawings]

FIG. 1 is a block diagram showing the configuration of an embodiment of a variable rate coding apparatus for image signals of the present invention.

FIG. 2 is a block diagram showing a configuration example of a decoding device corresponding to the encoding device of the present invention.

FIG. 3 is a block diagram showing a configuration example of a conventional variable rate coding device for image signals.

FIG. 4 is a diagram illustrating an example of a prediction process according to the MPEG method.

[Explanation of symbols]

 11 frame memory 12 buffer memory 13 coding rate control unit 21 buffer memory 22 inverse quantizer 23 frame control unit 31 quantization table 32 quantization control unit 33 quantizer 34 subtractor 35 adder 36 frame memory 37 buffer memory

Claims (2)

[Claims] 1. An intra-frame coded data or an inter-frame coded data is coded by switching each prediction mode of intra-frame coding, forward prediction coding, and bidirectional prediction coding at a predetermined use ratio. And a plurality of quantization tables used for the encoding process of the encoding means, and assigns a quantization table having a large quantization step size to the prediction error data by the bidirectional predictive encoding. , A quantization control means for receiving a coding rate instruction signal according to a congestion state of the network, selecting a corresponding quantization table, and setting a coding rate according to the quantization step size; Data, interframe predictive coded data, and a buffer memory that temporarily stores side information used for the prediction In the image signal variable rate coding device, when the coding rate instruction signal is received and when it instructs the reduction of the coding rate, the bidirectional prediction code is changed by changing the usage ratio of each prediction mode in the coding means. A variable rate coding apparatus for image signals, comprising a coding rate control means for performing control for increasing the ratio of coding. 2. An input signal is coded by switching each prediction mode of intraframe coding, forward prediction coding, and bidirectional prediction coding at a predetermined use ratio, and intraframe coded data or interframe coded data is converted. Encoding means to output, having a plurality of quantization tables to be used in the encoding process of the encoding means, select the corresponding quantization table by receiving the encoding rate instruction signal according to the congestion state of the network, Quantization control means for setting a coding rate according to the quantization step size, and a buffer memory for temporarily accumulating the intra-frame coded data, inter-frame predictive coded data and side information used for the prediction In the variable rate coding device for image signal, when the coding rate instruction signal is received and the instruction is to reduce the coding rate, A variable rate coding apparatus for image signals, comprising: a coding rate control means for sending only side information used for prediction of interframe predictive coded data from a buffer memory.