JP3170185B2 - Image decoding method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for transmitting a tool constituting an algorithm which is a means for decoding encoded information together with encoded information from an encoding apparatus, and reconstructing the tool received by the decoding apparatus as an algorithm. , An image decoding method for decoding received encoded information.

[0002]

2. Description of the Related Art In recent years, ISDN (Integrated Services)
With the spread of Digital Network (Service Integrated Digital Network), image communication services have been realized as new communication services. Videophone and videoconferencing systems are examples. Further, with the development of wireless transmission networks represented by PHS and FPLMTS, demands for further enhancement, diversification, and portability of services are rapidly increasing. In general,
When transmitting image information, such as a videophone or video conference system, the amount of image information is enormous, but in terms of the line speed and cost of the line used for transmission,
It is necessary to compress and encode the information amount of the image to be transmitted, and to transmit the image with the information amount reduced. JPE is used as a still image coding method as a coding method for compressing image information.
G (Joint Photographic coding Experts Group); H.261, MPEG1 (Moving Picture Coding Expert
Group) and MPEG2 have already been internationally standardized.
Further, MPEG4 standardization activity as an encoding method at an ultra-low bit rate of 64 kbps or less is being promoted.
According to MPEG4, since it is possible to flexibly cope with a variety of applications and to perform encoding in an optimal manner for each application, the existing JPEG, H.264, and MPEG4 standards are used. 26
1, not MPEG1 and MPEG2 encoding methods, which perform encoding in accordance with an algorithm, but each encoder tool (transformer, quantizer, inverse transformer, inverse quantizer, etc.)
It is necessary that a method of preparing a large number of codes and performing encoding by an appropriate combination of them.

[0005] FIG. FIG. 2 is a conceptual diagram illustrating the structure of a data string of encoded information obtained by encoding (compressing) image data in accordance with H.261. Each piece of coded information data (unsigned) such as motion vector information, DCT coefficients, and quantization steps shown in FIG. 3A is an image coded (compressed) based on a coding algorithm fixed in the coder. The information data, and the decoder decodes each of the received encoded information data with a decoding algorithm fixed corresponding to the encoding algorithm.

FIG. 5B is a conceptual diagram showing the structure of a data string of coded information obtained by coding (compressing) image data using a flexible coding method using an algorithm such as MPEG4. The data sequence of the coded information shown in FIG. 3B includes the coded (compressed) image information data such as the motion vector information 2, the variable coefficient 4, the motion vector information 6, the transform coefficient 8, the quantization step 10, and the like. , And tool information such as a motion compensation tool 1, an inverse transformation tool 3, a motion compensation tool 5, an inverse transformation tool 7, and a quantization tool 9 for decoding each of the image information data. In this case, each piece of tool information such as the motion compensation tool 1 can be selected from a plurality of types of tool information, and a combination of each piece of tool information can be freely selected. For this reason, the encoder transmits tool information used for encoding together with the image information data to the decoder, and the decoder decodes the tool information transmitted from the encoder when decoding the received image information data. It decodes the image information data encoded using the information.

As a method of implementing these encoding / decoding processes, a method of implementing dedicated hardware and software, and a method of executing appropriate software using a general-purpose arithmetic unit and a compiler. There is a way. First, a method of implementing the encoding process by mounting dedicated hardware and software will be described. FIG. FIG. 6 is a block diagram illustrating a configuration of an encoder that generates encoded information data illustrated in FIG. 6, an encoder includes an encoding control unit 11 for performing encoding control, a transform unit 12 for performing DCT transform, and a quantizing unit 1 for quantizing coefficients transformed by the transform unit 12.
3. It includes an inverse quantization unit 14 for performing inverse quantization of the coefficient quantized by the quantization unit 13, an inverse transformation unit 15 for performing inverse DCT transform, a memory 16, and a filter 17 in a loop.
Note that the memory 16 is a memory having a variable delay function for motion compensation used in motion-compensated inter-frame prediction, and the filter 17 is a filter in a loop that can be turned on / off for each macroblock.

When the encoding algorithm for generating the encoded information data shown in FIG. 5A is realized by dedicated hardware and software, the function of each tool constituting the algorithm is based on the encoding control shown in FIG. Unit 11, transformation unit 12, quantization unit 13, inverse quantization unit 14, inverse transformation unit 1
5. This is realized by dedicated hardware and software for the memory 16 having the motion compensation delay function and the loop filter 17, respectively. FIG. 261 is a block diagram illustrating a configuration of a decoder that decodes encoded information data encoded according to H.261. This decoder is configured by sharing the constituent elements of the encoder shown in FIG. 6, and the same components as those of the encoder shown in FIG. 6 are denoted by the same reference numerals. . That is, in FIG.
4 is an inverse quantization unit, 15 is an inverse transformation unit, 16 is a memory having a variable delay function for motion compensation, and 17 is a filter in a loop. The encoded information data encoded by the encoder shown in FIG. 6 is inversely quantized by the inverse quantization unit 14, and further, by the inverse transform unit 1.
In step 5, it is inverse DCT transformed and decoded. The memory 16
The in-loop filter 17 is used when decoding motion-compensated prediction encoded data.

As described above, H. In the case where several types of algorithms are processed using a method of performing encoding using a fixed algorithm such as H.261, hardware and software for realizing each algorithm are individually required. FIG. 1 is a block diagram illustrating a structure of an encoder that complies with H.261 and encodes a still image in accordance with JPEG. For example, with one terminal, a moving image is stored in H.264. When encoding according to J.261 and encoding a still image according to JPEG, the encoder has a configuration as shown in FIG. Thus, both the H.261 encoder 20 and the JPEG encoder 21 are provided independently. In FIG.
H. The H.261 encoder 20 and the JPEG encoder 21 receive moving image data and still image data, respectively, and output encoded information data that is compressed data.

When the algorithm for generating the encoded information data shown in FIG. 5B is realized by dedicated hardware and software, the encoder for realizing this algorithm is a circuit block of the encoder shown in FIG. 18 is realized by the configuration shown in FIG. That is, in this case, the encoder includes the conversion unit 12 and the quantization unit 1
3, each of the inverse quantization unit 14 and the inverse transformation unit 15 has a plurality of tools.
The necessary tools are selected from the converter tools A to X, the quantizer tools A to X, the inverse quantizer tools A to X, and the inverse transformer tools A to X shown in FIG.

A decoder for decoding the encoded information data shown in FIG. 5B is realized by replacing the circuit block 19 constituting the decoder shown in FIG. 7 with a circuit block 22 shown in FIG. Is done. That is, in this case,
The decoder has a plurality of types of tools of the inverse quantization unit 14 and the inverse transform unit 15, respectively, and the respective tools (inverse quantizer tools A to X and inverse transformer tools A to X shown in FIG. 9). And perform the decryption process by selecting the required tool. In this decoding process, the tool information of the motion compensation tool 1, the inverse transformation tool 3, the motion compensation tool 5, the inverse transformation tool 7, and the quantization tool 9 shown in FIG. 5B is transmitted to the control unit 23. The image information data of the motion vector information 2, the conversion coefficient 4, the motion vector information 6, and the conversion coefficient 8 following the respective tool information are transmitted to the respective tools for processing the respective image information data. At this time, the control unit 23 selects which tool (the inverse quantizer tools A to X and the inverse transformer tools A to X shown in FIG. 9) to use from each tool information, and Is processed and decoded by the tool selected by the control unit 23.

Next, a description will be given of a method of implementing a decoding process by executing appropriate software using a general-purpose arithmetic unit and a compiler. Hereinafter, the case of decoding the encoded information data having the structure shown in FIG. 5B will be described with reference to FIG. FIG. 10 is a block diagram illustrating a structure of a decoder including the general-purpose operation processing unit 24 and the compiler 25. The tool information of the motion compensation tool 1, the inverse transformation tool 3, the motion compensation tool 5, the transformation tool 7, the quantization tool 9 and the like shown in FIG.
5, the compiler 25 includes a general-purpose arithmetic processing unit 24
Generates a processing program for controlling the operation of. Also, motion vector information 2 following these tool information,
The image information data of the transform coefficient 4, the motion vector information 6, the transform coefficient 8, and the quantization step 10 are stored in the general-purpose arithmetic processing unit 2
4 given. Then, in accordance with the processing program generated by the compiler 25, the general-purpose operation processing unit 24
Processes the encoded image information data following the tool information, decodes the image information data, and generates decoded data.

Here, the tool information transmitted from the encoder includes not only the information indicating the type of the tool but also a tool body describing the processing procedure of the tool itself, so that the required tool can be transmitted to the decoding device side. , The received image information data can be decoded.

However, when transmitting the tool body and decoding the image information data using the transmitted tool body, if there is a transmission error, the received tool body may not operate normally. In particular, in the case of data transmission using a wireless transmission network, the error rate generally increases. If an error occurs in the data describing the processing procedure itself, such as the information data of the tool itself, not only the data processing of the system itself will not be executed properly, but in some cases, a fatal malfunction may be caused in the entire system .

In order to prevent such a situation, a method is generally adopted in which parity is added to transmission data to determine whether or not received data has an error. That is, a parity is added to a certain transmission block unit, and based on the result of calculating the parity and the data body, the receiving side determines whether or not there is an error in the received data. If the receiving side determines that there is no error, it returns success of reception, and if it determines that there is an error, it returns failure of reception. The transmitting side repeats the procedure of transmitting the next block when receiving a successful reception and retransmitting the current block when receiving a failed reception for each transmission block.

FIG. 11 is a flowchart for explaining an error determination operation algorithm in the conventional decoding apparatus, and is for a case where a data sequence having the structure shown in FIG. 5B is received and error determination is performed. is there. First, when the encoded information data is received, it is determined whether or not there is an error in the tool 1 included in the encoded information data (Step S).
1). If the result of this determination is that there is an error (step S1, No), a retransmission request for the tool 1 is issued to the transmitting side (step S2), and the retransmission data is waited for. If there is no error in the tool 1 (step S1, Yes), it is determined whether there is an error in the tool 2 (step S3). If the result of this determination is that there is an error (step S3, No), a request for retransmission of the tool 2 is issued to the transmitting side (step S4), and data retransmission is waited for. If there is no error in the tool 2 (step S3, Yes), a decoding process of the received encoded information data is executed using the tools 1 and 2 (step S3).
5) Wait for input of the next encoded information data.

When the next coded information data is received, it is determined whether there is any error in the tool 3 included in the coded information data (step S6). If the result of this determination is that there is an error (step S6, No), the tool 3
(Step S7), and waits for retransmission data. If there is no error in the tool 3 (step S6, Y
es), it is determined whether there is an error in the tool 4 (step S8). If the result of this determination is that there is an error (step S8, No), a retransmission request for the tool 4 is issued to the transmitting side (step S9), and the retransmission data is waited for. If there is no error in the tool 4 (step S8, Yes), it is determined whether there is an error in the tool 5 (step S10). If the result of this determination is that there is an error (step S10, N
o), a request for retransmission of the tool 5 is issued to the transmission side (step S)
11) Wait for retransmission data. If there is no error in the tool 5 (step 10, Yes), the tools 3, 4 and 5
Is used to execute decoding processing of the received encoded information data (step S12), and waits for input of the next encoded information data.

[0016]

According to the above-described error determination algorithm, if there is at least one tool in which an error is detected, even if other tools are normal, decoding using a normal tool is performed. No processing is performed and no decoding is performed until no errors are detected by all tools. Further, for example, when an error is detected in a certain transmission block in the tool 1 included in the received encoded information data (step S1, No), the retransmission procedure of the tool 1 (step S2) is completed and the retransmission is performed. The error is determined again for the tool (step S1), and the next transmission block (tool 2) is not transmitted from the encoder side until it is determined that the tool 1 has been normally received. For this reason, when the number of retransmission procedures increases, there is a problem that the operating efficiency of the receiving side is remarkably reduced, and the use efficiency of the communication transmission path is reduced. Furthermore, even if the transmission of the tool is performed normally, there is a problem that whether the tool normally operates on the receiving side cannot be determined by the transmitting terminal and the receiving terminal.

The present invention has been made in view of such a problem, and it is determined on the receiving side (decoder side) whether or not the decoding tool and algorithm operate normally, and the communication transmission path and the receiving side are determined. object to provide an image decoding method capable of retransmitting without lowering the operation efficiency of the side of the terminal device, a decoding tools and algorithms determined not to work properly at the decoder side from the transmitting side And

[0018]

Means for Solving the Problems The present invention has the following arrangement to achieve the above object. According to the present invention
The image decoding method is a method for decoding an image from an image encoding device.
The tools that make up the algorithm and those of each of the tools
A code consisting of each image information data decoded by each
Received the information string of the conversion information, using each of the tools received
The image information data to be processed by each tool
Image decoding method for restoring an image by decoding
Whether the received tool operates normally
Is determined independently for each tool, and the
If it is determined that the operation always occurs, immediately after the determination, the tool
Image information data to be processed by the tool using the
Specially executes the decryption process of each tool independently.
Sign.

The image decoding method according to the present invention is characterized in that
If one of the files is determined to not work properly,
Immediately thereafter, the retransmission of the tool is transmitted to the image encoding device.
And the re-transmission procedure of the tool is performed by the image coding apparatus.
Whether the following tools work properly while running
Is determined .

The image decoding method according to the present invention is characterized in that
The judgment as to whether or not the tool that has been operated normally
Based on the test routine sent from the encoding device,
It is specially performed by performing the operation wipe test of each tool.
Sign.

[0021]

[0022]

[0023]

DESCRIPTION OF THE PREFERRED EMBODIMENTS (First Embodiment) An embodiment of the present invention will be described below with reference to the drawings. FIG. 1 is a diagram showing an MPEG4 image according to the image decoding method of the present invention.
FIG. 3 is a conceptual diagram illustrating an example of a configuration of an encoded information data sequence according to a flexible encoding method such as that shown in FIG. The encoded information data shown in the figure includes a test routine 31 which is a program routine for performing an operation test of the decoding algorithm,
Subsequent motion compensation tool 32a, inverse conversion tool 32
b, tool information such as an inverse quantization tool 32c, motion vector information 33a, a transform coefficient 33b, a quantization step 33
c and other image information data. The test routine 31 includes test items for each tool, and performs an operation test of the tool information 32a to 32c constituting the decoding algorithm based on the test items, thereby forming the test information from the tool information 32a to 32c. The operation test of the decoding algorithm to be performed can be performed on the decoder side.

FIG. 2 is a flowchart for explaining an algorithm for decoding encoded information data having the structure shown in FIG. First, when the reception of all of the tool information 32a to 32c constituting the decoding algorithm is completed, each of the tools 32a to 32c is set based on the test routine 31.
The operation test of 32c is sequentially performed, and the result of the operation test is determined each time. Details will be described below. First, a decoding algorithm test is executed based on the test routine 31 included in the received encoded information data (step S21). Next, the operation shifts to the operation test of each tool, and the operation test of the tool 32a is executed and determined (Step S2).
2). If it is determined in this determination that the operation of the tool 32a is normal (step S22, Yes), the tool 32a
The decoding of the motion vector information 33a by a is executed (step S24), and the operation shifts to the operation test of the tool 32b. If it is determined in this determination that the operation of the tool 32a is abnormal (step S22, No), a request for retransmission of the tool 32a is made to the transmission side, and then the operation immediately proceeds to the operation test of the tool 32b. .

Next, an operation test of the tool 32b is executed to make a determination (step S25). In this determination, the tool 32
b is determined to be normal (step S2
5, Yes), the conversion coefficient information 33b by the tool 32b
Is executed (step S27), and the tool 32c
Move to the operation test. Also, in this determination, the tool 32
If it is determined that the operation of “b” is abnormal (No at Step S25), the operation immediately proceeds to the operation test of the tool 32c after requesting the transmission side to retransmit the tool 32b. Next, an operation test of the tool 32c is executed to make a determination (step S28). If it is determined that the operation of the tool 32c is normal (step S28, Yes), the quantization step information 33c is decoded by the tool 32c (step S30), and the next encoded information data is decoded. Wait for reception. If it is determined in this determination that the operation of the tool 32c is abnormal (step S28, N
o) After requesting the transmission side to retransmit the tool 32c, wait for reception of the next encoded information data.

As a result of the above-mentioned tool operation test, a retransmission request is made to the encoder side for a tool determined not to operate normally (steps S23, S26, S29).
Then, it immediately shifts to the operation test of the next tool. When the operation test is completed normally, decoding of the image information data to be processed by the tool is started (step S2).
4, 27, 30). For example, when the operation test result of the tool 32a is determined to be abnormal, the decoder requests the encoder to retransmit the tool 32a (step S23). While the retransmission procedure of the tool 32a is executed on the encoder side, an operation test of the tool 32b is executed on the decoder side, and the test result is determined (step S25).

On the other hand, in the conventional example shown in FIG.
When the decoder side detects an error in the transmission block of the tool 1, the encoder side executes a retransmission procedure for this transmission block (step S2), but the next block is transmitted until the retransmission procedure is completed. Therefore, when a transmission path with many errors is used, the processing efficiency on the decoder side decreases, and the start of the decoding process of the encoded information data received on the decoder side is delayed. As described above, in the first embodiment of the present invention, it is possible to perform the operation test of each tool on the decoder side by transmitting the test routine for performing the operation test of the tool together with the tool. Becomes Further, while the retransmission procedure of the tool determined to be abnormal in operation as a result of the operation test is being executed on the encoder side, the operation test of the next tool is executed on the decoder side, so that communication with the decoder is performed. The operation efficiency of the transmission path can be increased.

(Second Embodiment) FIG. 3 is a conceptual diagram showing another example of the configuration of an encoded information data sequence according to a flexible encoding method such as MPEG4 by the image decoding method of the present invention. Each tool 42a-4
2c is individually provided with test routines 41a to 41c, respectively, so that an operation test for each tool can be independently performed by a decoder on the receiving side. FIG. 4 is a flowchart for explaining an algorithm for decoding the encoded information data having the structure shown in FIG.
Upon receiving the tool information included in the encoded information data,
Immediately, based on the test routine provided in the received tool information, an operation test of the tool is executed to determine the operation (steps S31, S35, S39). As a result of this determination, when the operation is determined to be abnormal, a request is made to the encoder side to retransmit the tool (steps S33, S37, S4).
1) If the operation is determined to be normal, the tool is used to execute decoding of data to which processing is assigned to the tool (steps S34, S38, S42). That is, in the present embodiment, the operation test is executed for each tool, and the retransmission procedure is executed for each tool.

The details will be described below. First, an operation test of the tool 42a included in the received encoded information data is performed, and the test result is determined (step S31). If the result of this determination is that the operation is normal, a decryption process is performed using this tool (step S34), and the operation proceeds to the operation test of the tool 42b. Further, when it is determined that there is an abnormality, after requesting retransmission of the tool 42a to the transmitting side (step S33), the operation immediately proceeds to the operation test of the tool 42b. Hereinafter, similarly, an operation test of the tools 42b and 42c is performed, and based on the test result, a request is made to the transmitting side to retransmit the corresponding tool. Tools that work properly are used immediately for the decryption process. After a series of operation tests of the tool and the decoding process are performed, an operation test of the retransmitted tool is performed. In the re-operation test, a tool determined not to operate normally requests retransmission again, while a tool determined to operate normally performs decoding processing using the request. As long as there is a tool whose operation is determined to be abnormal in the operation test, the decoder side
Request retransmission repeatedly. Further, while the retransmission procedure is being executed on the encoder side, decoding processing using a tool whose operation has been determined to be normal is executed on the decoder side.
As described above, in the second embodiment, as in the first embodiment, while the retransmission procedure of the tool determined to be abnormal in operation is being executed on the encoder side, the decoder side Then, the operation test of the next tool can be executed, so that the operation efficiency of the decoder and the communication transmission line can be improved, and further, the operation test is performed for each tool by providing a test routine for each tool. be able to.

[0030]

As apparent from the above description, the present invention has the following effects. (1) When transmitting a tool constituting an algorithm which is a means for decoding encoded information data from the encoder side,
By simultaneously transmitting a test routine for performing an operation test of the algorithm, it becomes possible for the decoder to determine whether the received decoding algorithm normally operates on the decoder side. (2) Since the test routine of the decoding algorithm includes a test item for each tool constituting the decoding algorithm, it is determined whether or not each decoder constituting the decoding algorithm normally operates on the decoder side. Judgment becomes possible. (3) By providing a test routine for each tool constituting the decoding algorithm, an operation test can be performed independently for each tool, and whether or not the tool received at the decoder operates normally on the decoder side Can be determined independently for each tool. (4) By retransmitting the tool determined to be abnormal in operation, the decoder can obtain a normal tool, and the decoder can use this tool to prevent errors in decoding processing. Becomes (5) As a result of the operation test on the decoder side, with respect to the tool which operates normally, the tool immediately starts the decoding process of the encoded information data (image information data) to be processed by the tool, thereby communicating with the decoder. The operation efficiency of the transmission path can be improved.

[Brief description of the drawings]

FIG. 1 is a conceptual diagram illustrating a configuration of a transmission data sequence according to an image decoding method according to a first embodiment of the present invention.

FIG. 2 is a flowchart illustrating an operation algorithm of a decoder according to the image decoding method according to the first embodiment of the present invention.

FIG. 3 is a conceptual diagram illustrating a configuration of a transmission data sequence according to an image decoding method according to a second embodiment of the present invention.

FIG. 4 is a flowchart illustrating an operation algorithm of a decoder according to the image decoding method according to the second embodiment of the present invention.

FIG. FIG. 2 is a conceptual diagram illustrating the structure of a data string of encoded information obtained by encoding (compressing) image data in accordance with H.261. FIG. 2B is a conceptual diagram illustrating a structure of a data string of encoded information obtained by encoding (compressing) image data using a flexible encoding method such as an MPEG4 algorithm.

FIG. 2 is a block diagram illustrating a configuration of an encoder conforming to H.261.

FIG. 261 is a block diagram illustrating a configuration of a decoder conforming to H.261.

FIG. 8 is a block diagram illustrating a configuration of an encoder including a plurality of algorithms.

FIG. 9 is a block diagram illustrating a configuration of a decoder implemented using dedicated hardware and software.

FIG. 10 is a block diagram illustrating a configuration of a decoder implemented using a general-purpose operation processing unit and a compiler.

FIG. 11 is a flowchart illustrating an operation algorithm of a decoder.

[Explanation of symbols]

 31, 41a to 41c Test routine 32a, 42a Motion compensation tool 32b, 42b Inverse transformation tool 32c, 42c Inverse quantization tool 33a, 43a Motion vector information 33b, 43b Transform coefficient 33c, 43c Quantization step

Continuation of the front page (58) Field surveyed (Int.Cl. ⁷ , DB name) G06F 13/00 G06F 9/445 G06F 11/00

Claims (3)

(57) [Claims] 1. An image decoding apparatus, comprising:
Each tool that composes the algorithm and each of the above tools
Encoding consisting of each image information data decoded by
Receiving the information data sequence, using each of the received tools
Each of the image information data to be processed by each tool
To restore the image by decoding
In, the determination whether the received tool operates normally,
This is performed independently for each tool so that the tool operates normally.
If it is determined that the tool is used immediately after the determination
Decoding of image information data to be processed by the tool
Is executed independently for each tool.
Image decoding method. 2. The image decoding method according to claim 1, wherein
And it is determined that one of the tools does not work properly
Immediately after the determination, the retransmission of the tool
Device, and the image encoding device
While the resend procedure is running, the following tools work properly:
Image decoding method for determining whether or not to make
Law. 3. An image decoding method according to claim 1, wherein
In the method, the determination as to whether the received tool operates normally,
Based on a test routine transmitted from the image encoding device,
Based on the wiping test of each tool
An image decoding method, characterized in that: