WO2003036988A1

WO2003036988A1 - Media separating method, image decoding method and image decoding device

Info

Publication number: WO2003036988A1
Application number: PCT/JP2001/009368
Authority: WO
Inventors: Koji Imura
Original assignee: Matsushita Electric Industrial Co., Ltd.
Priority date: 2001-10-25
Filing date: 2001-10-25
Publication date: 2003-05-01

Abstract

In media separation, when an error is detected in an AL-SDU image packet, the AL-SDU image packet is throw away, an identification code identifiable by an image decoding unit (107) is inserted in place, and an error position is notified to the image decoding unit (107) in units of a packet. When the image decoding unit (107) detects the identification code in a received image signal, the image decoded immediately before an error is detected is displayed, and the image between immediately after the error is detected and before a next synchronization word is detected is concealed.

Description

TECHNICAL FIELD The present invention relates to a media separation method and an image decoding method and apparatus.

The present invention relates to a media separation method, an image decoding method, and a device suitable for use in a communication device for performing moving image communication such as a videophone.冃.

Fine 1

Conventionally, as a technology for encoding video and audio signals and multiplexing them into one signal, Reco-Viewing Endation H recommended by the International Telecommunication Union (ITU-T: International 'Telecommunication Union)) There is .223. This H.223 includes an adaptation layer (three types from AL1 to AL3) that packetizes with an appropriate layer according to the characteristics of the media to be transmitted (audio and video are called media). There is a max layer that combines and multiplexes the packets output from the adaptation layer.

The appropriate adaptation layer is selected from AL1 to AL3 according to the characteristics of the media, but AL3 is generally selected for a moving image signal. AL3 is said to be suitable for real-time image communication because it has an error detection function and a retransmission control function. A packet that has been bucketed by the adaptation layer is called an AL-SDU (adaptation layer, service, data unit) packet. The max layer collects the AL-SDU packets output from AL1 to AL3 into one transmission data stream. The multiplexing process is performed in this max layer.

At the time of reception, packets are separated for each medium in the max layer, and the separated buckets are input to the corresponding adaptation layers. In this case, if it is a moving image signal, the AL-SDU packet is input to AL3 and error detection processing is performed. The data after error detection processing is sent to an image decoder (not shown). ). If an error is detected by the error detection process, whether to retransmit the AL-SDU packet containing the error and whether to pass the AL-SDU packet to the image decoder (not shown) are as follows. Since the content is outside the recommendations of -Τ, it can be freely selected.

As a technique for encoding moving picture signals, there is IS0 / IEC 14496 part 2 (Visual) (commonly known as MPEG-4) recommended by the International Standards Committee (IS0 / IEC). This MPEG-4 is a multimedia encoding method, and is a technology that can encode various video materials. This image coding technology is realized by three technologies: “motion-compensated predictive coding”, “discrete cosine transform”, and “variable-length code”.

In motion-compensated predictive coding, the input image is compared with the previous coded image, and the amount of motion between them is measured (motion detection). Next, the input image is predicted based on the measured amount of motion and the previous encoded image. Then, the difference (prediction error signal) between the predicted image (predicted image) and the input image is calculated. Next, the calculated prediction error signal and the motion amount described above are transmitted to the receiving side. By adopting this method, image information transmission with a small data amount becomes possible. The discrete cosine transform is for transforming the above-mentioned prediction error signal into the frequency domain. When the prediction error signal is converted to the frequency domain, the power is concentrated in a specific frequency domain (low frequency domain). By combining this feature with variable-length coding, images with a smaller amount of data can be obtained. Information transmission becomes possible. In variable-length coding, when the frequency of occurrence of data to be encoded is biased, the bias is used to shorten the code length for events with a high frequency of occurrence, and for events with a low frequency of occurrence. In this method, the average code length is shortened by expressing it with a long code length. These three techniques, “motion-compensated predictive coding”, “discrete cosine transform”, and “variable-length coding”, do not apply to the entire image, but block one image into a 16 × 16 pixel block (encoding). This is applied to each coding block divided into blocks.

By the way, in recent years, with the spread of wireless communication terminals such as mobile phones, attention has been paid to wireless transmission technology for moving image signals. Wireless transmission requires a coding technique that has transmission error resistance because the transmission error rate is higher than that of wired transmission. Transmission error A technique called a video bucket is known as a technique for suppressing the image quality degradation due to image quality. In this method, data for one screen or one or more encoded blocks is transmitted as one transmission unit (video packet).

As an example of a video packet configuration method, there is a method of accumulating a generated code amount for each coding block and forming a video packet when a certain code amount is reached (for example, Japanese Patent Application Laid-Open No. No. 14). By adopting this method, the video packet is composed of many coded blocks in the area where the amount of code is small, such as the background, and the video packet is formed by the small coding block in the area where the motion is large and the amount of code is large. Be composed. As an advantage, even when an error occurs in a video packet due to a transmission error, the portion with large motion is composed of few coded blocks, so that the range of deterioration can be suppressed to a small value. In addition, the background part with small movement contains many coding blocks, but the deterioration is not conspicuous because there is almost no movement. By employing this method, bits can be efficiently allocated.

Also, a technique called concealment processing is known as a technique for suppressing image quality degradation due to transmission errors. This is a technique that replaces the coded block portion lost due to a transmission error with other information, thereby compensating for the image quality degradation of the lost portion. An example of this concealment processing technology is shown as an important but important technology outside the MPEG-4 recommendation.

On the other hand, as a technique capable of detecting an error in encoded data, there is a technique disclosed in, for example, Japanese Patent Application Laid-Open No. 7-321750.

FIG. 1 is a block diagram showing a configuration of the moving image communication device disclosed in the publication. In this figure, a transmitting device 1 includes an encoder 2 for encoding a signal to be transmitted, a specific code sequence detector 3 for detecting a specific code sequence from an encoded data sequence, and a specific code. A communication control unit 4 that controls transmission of encoded data including a specific code sequence to the receiving device 5 as a transmission unit each time a sequence is detected. The receiving device 5 includes a communication control unit 6 for receiving the encoded data transmitted from the transmitting device 1, an error detector 7 for determining whether the received data is correct, The decoder 9 is inserted between the communication controller 6 and the encoder 9 for decoding the data, and generates a new data stream based on the error notification signal from the error detector 7 and A code replacement unit 8 is provided which replaces a part of the message and sends it to a decoder 9.

In the transmitting side device 1, the data sequentially generated in time series from the encoder 2 is written into the transmission buffer of the communication control unit 4 and, at the same time, is sequentially inspected by the specific code sequence detector 3. When specific time-series data is generated from the encoder 2, the specific code sequence detector 3 notifies the communication control unit 4 of the detection of the specific code. Upon receiving this notification, the communication control unit 4 switches the transmission buffer to which the data is written, continuously writes the detected specific code and the data thereafter to the transmission buffer of the switching destination, and further writes the previously written transmission buffer. Sends the contents of the buffer to communication line 10. Here, the specific code to be detected by the specific code sequence detector 3 is a code (start code) indicating the beginning of the unit data (encoded data) from the encoder 2.

On the other hand, in the receiving device 5, the communication control unit 6 writes the received data into the receiving buffer, and when all the data sent from the transmitting device 1 has been received, switches the receiving buffer and switches. The data of the next receiving buffer sent from the transmitting device 1 is written to the previous receiving buffer. The error detector 7 checks whether there is an error in the received data. If there is an error, the error detector 7 notifies the code replacement unit 8 of the error detection. The code replacement unit 8 reads the content from the reception buffer in which the communication control unit 6 has finished writing the received data and sends it to the decoder 9, but when receiving an error detection notification from the error detector 7, The data in the reception buffer is discarded, a code indicating a data error is newly generated, replaced with the position of the specific code, and sent to the encoder 9.

FIG. 2 is a diagram illustrating a data sequence output from the encoder 2, which is a time sequence from left to right. The coded data A32, coded data B33, and coded data C36 are output in this order, and each coded data has a start code A31, a start code B33, and a start code indicating the beginning. C 35 is added. These start codes A 31, B 33, and C 35 are detected as specific codes by the specific code string detection unit 12. Code.

On the other hand, FIG. 3 is a diagram showing a sequence of data input to the decoder 9, and shows a time sequence from left to right. Encoded data D42 is input following the start code D41. If there is an error in the next error, an error code 43 is input, and then an error message 44 is input.

As described above, when a communication error occurs, the start code is replaced with the error code in the receiving device 5, so that the decoder 9 processes erroneous data at an unpredictable position in the encoded data. To prevent the display of large image disturbances.

However, the conventional technology has the following problems.

(I) Problems in media separation processing

There was no provision in the adaptation layer AL3 for the handling of an AL-SDU bucket where an error was detected.

The following methods can be considered for handling AL-SDU packets with mixed errors.

(1) The AL-SDU packet in which the error was detected is passed to the image decoder as it is.

(2) Discard the AL-SDU packet in which the error was detected.

(3) Store the bit position where the error occurred.

However, these methods have the following problems.

In the method (1), the error detection processing of the adaptation layer AL3 is completely useless.

In the method (2), the packets before and after the AL-SDU packet in which the error was detected are connected, and the AL-SDU packets before and after have no errors, but inconsistency occurs at the boundary.

In the method of (3), it is necessary to manage the memory for storing the bit position where the error has occurred and the number of bits and bits. If transmission errors frequently occur, the memory capacity to be theoretically secured cannot be determined.

(II) Problems when transmission errors occur in image coding technology (1) If a transmission error occurs in a bit string using a variable-length code, the decoder cannot identify the position where the transmission error occurred.

The above-mentioned variable-length codes have different code lengths in order to increase coding efficiency. However, if a transmission error occurs, the code delimiter position shifts, and data that differs from the original data from the encoder is lost. It may be decoded as evening. Also, due to a transmission error, a code that does not exist in the variable length code may not be able to be decoded. As described above, if a transmission error occurs once, decoding is continued with the delimiter position of the code shifted, and very often a code that cannot be decoded is detected during that time. If there is, the error location cannot be specified.

(2) In order to prevent display of a distorted image when a transmission error is detected, concealment is performed such that a coded block having a transmission error is not displayed, and instead, for example, an image at the same position as the previous screen is replaced. Processing may be performed. In this case, in order to perform the concealment processing, it is necessary to determine which coding block to which coding block is to be targeted. However, as described above, even if a transmission error can be detected, the error position cannot be specified, so that the coding block that may have been correctly decoded is discarded. Performs either concealment processing in units of video packets including AL-SDU packets or concealment processing from coding blocks that have detected errors on the assumption that erroneously decoded images are displayed. It is common. As described above, since an error position cannot be specified, an appropriate concealment processing range cannot be determined, and a problem that causes image quality deterioration is caused.

(III) Problems in the moving picture communication device disclosed in Japanese Patent Application Publication No. 7-311750

When the receiving apparatus 5 detects an error in the received encoded data, it outputs or discards the error data as error data, and discards the start code at the head of the data to generate and output an error code. As a result, it is possible to easily realize processing that minimizes quality degradation in the decoding processing. However Therefore, the coded data itself having an error is output as an error message or discarded, so that the quality deterioration cannot always be sufficiently suppressed. For example, if the coded data is one line of multiple lines out of all the lines of one screen, discarding them altogether makes it difficult to reproduce the corresponding image part. In addition, since there is no means for specifying an error part even if it is treated as an error (no means for specifying an error part is described), accurate image reproduction cannot be expected. Disclosure of the invention

An object of the present invention is to provide a media separation method, an image decoding method, and an apparatus capable of minimizing image quality degradation when a transmission error occurs. The purpose of this is to discard the AL-SDU bucket if there is a transmission error in the AL-SDU packet separated from the multiplexed signal, and insert a specific code instead to notify that there is an error in the image decoding process. Is achieved by The specific code does not exist in the variable-length code table and can be specified in the image decoding process. BRIEF DESCRIPTION OF THE FIGURES

FIG. 1 is a block diagram showing a configuration of a conventional moving image communication device;

FIG. 2 is a diagram for explaining the operation of the conventional moving image communication device;

FIG. 3 is a diagram for explaining the operation of the conventional moving image communication device;

FIG. 4 is a block diagram showing a configuration of a receiving side of the image communication terminal device according to Embodiment 1 of the present invention;

FIG. 5 is a block diagram showing a configuration of an image decoding unit of the image communication terminal device according to Embodiment 1 of the present invention;

FIG. 6 is a diagram for explaining the operation of the image communication terminal device according to Embodiment 1 of the present invention;

FIG. 7 is a block diagram showing a configuration of an image decoding unit of the image communication terminal device according to Embodiment 2 of the present invention; FIG. 8 is a diagram for explaining the operation of the image decoding unit of the image communication terminal device according to Embodiment 2 of the present invention;

FIG. 9 is a block diagram showing a configuration of an image decoding unit of the image communication terminal device according to Embodiment 3 of the present invention;

FIG. 10 is a diagram for explaining an operation of the image decoding unit of the image communication terminal device according to Embodiment 3 of the present invention. BEST MODE FOR CARRYING OUT THE INVENTION

Hereinafter, the best mode for carrying out the invention will be described in detail with reference to the drawings.

(Embodiment 1)

FIG. 4 is a block diagram showing a configuration on the receiving side of the image communication terminal device according to Embodiment 1 of the present invention.

In FIG. 4, the image communication terminal apparatus according to the present embodiment includes a receiving antenna 101, a radio receiving section 102 for receiving a radio signal via receiving antenna 101, and a radio receiving section 102. Demodulation unit 103 for demodulating the multiplexed signal from the received signal output from the unit, and a media separation unit (media separation device) for separating the multiplexed signal demodulated by the demodulation unit 103 into an image signal and an audio signal. 104, ROM 105 storing a program for performing media separation processing, RAM (work memory) 106 used in media separation processing, and media separation unit 1 An image decoding unit 107 for restoring the original image signal from the AL-SDU image bucket in the image signal separated at 04, a ROM 108 storing a program for performing image decoding processing, RAM (work memory) used in image decoding processing The D / A conversion section 110 converts the image signal into analog, the display section 111 displays the image visually, and the AL-SDU audio packet in the audio signal separated by the media separation section 110. It comprises a voice decoding unit 112 for restoring the audio signal, a D / A conversion unit 113 for converting the audio signal into an analog signal, and a speaker 114 for outputting audio. The media separation unit 104 and the image decoding unit 107 are examples. For example, it is realized by a DSP (Digital Signal Processor).

FIG. 5 is a block diagram showing the configuration of the media separation unit 104. As shown in this figure, the media demultiplexing unit 104 includes a synchronizing word detector 201 for detecting a segment (synchronizing word) of an AL-SDU image packet from the multiplexed signal, and an AL signal from the multiplexed signal. -Media separator 202 for separating S-SDU image packets and AL-SDU voice packets, Error detector 203 for detecting AL-SDU image packets with transmission errors, and specific code generation for generating specific codes And a switching switch 205. When detecting the synchronization word from the multiplexed signal demodulated by the demodulation unit 103, the synchronization word detector 201 notifies the media separator 202 of the detection. Upon receiving notification from the sync word detector 201 that the sync word has been detected, the media separator 202 separates the AL-SDU image packet and the AL-SDU voice packet from the multiplexed signal at that time. To Here, the portion sandwiched between the synchronization words becomes one multiplexed packet, and synchronization is established.

The separated AL-SDU image packet is subjected to error detection processing by the error detector 203. When the error detector 203 detects an erroneous AL-SDU image packet, it switches the switch 205 from the media separator 202 to the specific code generator 204 and outputs the erroneous AL. -Do not output the SDU image packet, and output the specific code generated by the specific code generator 204 instead. In this case, the specific code generated by the specific code generator 204 can be completely distinguished from the variable length code used in the image decoding unit 107 (see FIG. 1). That is, it is a code that does not exist in the variable-length code table including various codes for representing an image and can be recognized by the image decoding unit 107. If the error detector 203 does not detect an erroneous AL-SDU image packet, the switching switch 205 maintains the state of switching to the media separator 202 side. The AL-SDU image packet separated in 202 is output as it is. The error detector 203 switches the switching switch 205 to the specific code generator 204 side, and then sends it to the media separator 202 at the timing of the synchronization word of the multiplex signal transmitted next. Switch. Here, FIG. 6 shows the process of creating / separating a multiplexed signal and assigning a specific code.

First, a block is generated by dividing one image into, for example, 16 × 16. Next, each of the generated blocks is encoded. After encoding all blocks for one image, a video packet is generated using, for example, encoding blocks # 1 to # 4 as one transmission unit. Then, an AL-SDU image packet is configured by adding an error detection code or the like to each transmission unit. In this case, if one transmission unit is long, it may be divided into multiple AL-SDU image packets. Figure 6 shows a case where the data is divided into two. After generating the AL-SDU image packet, multiplex it with the AL-SDU voice packet and transmit it as a multiplexed signal. Then, when the receiving side receives the transmitted multiplexed signal, it separates the AL-SDU packet of the image and the AL-SDU packet of the audio based on the header information of the multiplexed signal. If there is an error in the separated AL-SDU image packet, the AL-SDU image packet is discarded and a specific code is inserted instead. When a specific code is inserted, decoding is not performed until the next transmission unit.

Next, an operation of the image communication terminal apparatus according to the present embodiment will be described with reference to FIGS.

A radio signal captured by antenna 101 is received by radio reception section 102, and a multiplexed signal is demodulated by demodulation section 103. The demodulated multiplexed signal is separated into an AL-SDU image packet and an AL-SDU audio packet by a media separation unit 104. At this time, the separated AL-SDU image packet and AL-SDU audio packet are temporarily stored in the RAM 106. Then, the separated AL-SDU image packet is input to the image decoding unit 107 and decoded into a digital image signal, and then converted to an analog image signal by the D / A conversion unit 110. This is visually displayed on the display unit 111. If an erroneous AL-SDU image bucket is detected in the multiplexed signal sent from the transmitting side (not shown), the packet is not output and the corresponding image is displayed on the display unit. 1 Not displayed on 1. On the other hand, the separated AL-SDU voice packet is input to the voice decoding unit 112, where it is decoded into a digital voice signal, converted to an analog voice signal by the DZA conversion unit 113, and transmitted from the speaker 114. Voice output. As described above, according to the image communication terminal apparatus according to the present embodiment, an image signal for one image is divided into a plurality of blocks to generate blocks, and after each block is encoded, the plurality is divided into one transmission unit. Then, an error is detected for each AL-SDU image packet generated by dividing or dividing this video bucket as it is, and for the AL-SDU image packet having an error, the image decoding unit 107 Since it is replaced with a specific code that can be specified, it is possible to minimize image quality degradation when a transmission error occurs. In particular, according to the present invention, since an error is detected in AL-SDU image packet units, the moving image disclosed in Japanese Patent Application Laid-Open No. 7-321750 does not detect errors in AL-SDU image packet units. It is possible to suppress image quality degradation at the time of transmission error occurrence lower than in the image communication device.

(Embodiment 2)

FIG. 7 is a block diagram showing a configuration of the image decoding unit 107 of the image communication terminal device according to Embodiment 2 of the present invention. Note that portions other than the image decoding unit 107 are the same as those of the image communication terminal device according to Embodiment 1 described above, and therefore, FIG. 1 will be referred to when necessary for the description.

In FIG. 7, the image decoding unit 107 of the image communication terminal device according to the present embodiment includes a variable-length decoder 301, an inverse quantizer 302, and an inverse discrete cosine transformer 303. The frame memory 304, the read from the frame memory 304, the Z address generator that generates the Z write address, the output of the discrete cosine transformer 303, and the read from the frame memory 304 It comprises an adder 306 for adding data, a specific code detector 307 for detecting a specific code, and a switching switch 308.

The variable-length decoder 301 decodes a motion vector such as the position of a coded block in an image and how much the coded block has moved with respect to the previous frame, and calculates the difference between the images in frequency. Decode the converted signal (hereinafter, frequency signal). The inverse quantizer 302 performs a process of inversely quantizing the frequency signal of the image with a predetermined value to return to a frequency signal. The inverse discrete cosine transformer 303 performs a process of converting a frequency signal into an image signal. Note that this image signal is a prediction error signal for the previous frame. No. The address generator 305 generates a read address based on the coded block address decoded by the variable length decoder 301 and the motion vector to cut out the area of the previous frame and read it. The image signal read by the read address is a predicted image.

The adder 303 adds the prediction error signal and the prediction image to form a reproduced image. The specific code detector 307 detects a specific code from the received image signal. As long as the specific code is not detected, the state in which the switching switch 308 is switched to the adder 306 is maintained, and the decoded block data that has been decoded is output. When a specific code is detected, the changeover switch 308 is switched from the adder 306 to the “0” data side to stop outputting the coded block data. The output coded block data is written to the frame memory 304 for use as a predicted image of the next frame.

FIG. 8 is a diagram showing an example of the operation of the image communication terminal device according to the present embodiment.For example, when a specific code is detected at the position shown in the drawing of coding block # 3, the coding block immediately before that is detected. That is, the position of the encoding block # 2 is specified, and it is determined that the encoding blocks prior to the encoding block # 2, that is, the encoding blocks # 2 and # 1, have been correctly recognized, and a display is made for these. Note that in this example, the video bucket is divided into three AL-SDU image packets, and an error detection code is added to each of them. .

As described above, according to the image communication terminal apparatus according to the present embodiment, when a specific code is detected in an image signal, an error-free encoded program up to immediately before the detection of the specific code is decoded. Since the display is performed, it is possible to further suppress the deterioration of the image quality as compared with the image communication terminal device according to the first embodiment.

(Embodiment 3)

FIG. 9 is a block diagram showing a configuration of the image decoding unit 117 of the image communication terminal device according to Embodiment 3 of the present invention. Note that, in this figure, the same parts as those in FIG. 7 described above are denoted by the same reference numerals. Also, in the present embodiment, image decoding Since the components on the receiving side other than the components 117 are the same as those of the image communication terminal device according to the above-described first embodiment, FIG. 1 will be referred to when necessary for the description. In FIG. 9, the image decoding unit 1 17 of the image communication terminal apparatus according to the present embodiment includes a variable-length decoder 301, an inverse quantizer 302, and an inverse discrete cosine transformer 303. , A frame memory 304, an address generator 305, an adder 306, a special code detector 307, a switch 308, and a synchronous word detector 309. It is composed.

The image signal separated by the media separation unit 104 (see FIG. 4) is input to the variable-length decoder 301 so that the position of the coding block in the image and the position of the block are different from those of the previous frame. The motion vector, such as how much the camera has moved, is decoded, and the frequency signal of the image is decoded. The frequency signal of the image is inversely quantized by the inverse quantizer 302 and returned to the frequency signal. Then, the returned frequency signal is converted into an image signal by the inverse discrete cosine transformer 303. This image signal is a prediction error signal for the previous frame. Also, based on the coded block address and the motion vector decoded by the variable-length decoder 301, a read address from which the area of the previous frame is cut out and read is output from the address generator 305. Then, an image signal corresponding to the read address is read from the frame memory 304. The read image signal is a predicted image.

On the other hand, when the specific code detector 307 detects a specific code from the image signal separated by the media separation unit 104 and inputs it to the address generator 305, the address generator 05 discards the generated address and outputs the address calculated only from the encoded program address. As a result, data at the same position as the current block to be processed in the previous frame is read from the frame memory 304.

When the synchronization word detector 309 detects a synchronization word indicating the start of an image packet / frame from the image signals separated by the media separation unit 104, the address generator 305 Outputs the address generated from the motion vector and the coded block address. In this case, the specific code is detected by the specific code detector 307. If the address calculated only from the coded block address is output, the output is changed to output the motion vector and the address generated from the coded block address.

The address generator 305 maintains the current output state unless the specific code detector 307 and the synchronizing word detector 309 detect anything. When the specific code is detected by the specific code detection unit 307, the switching switch 308 is switched to “0” to the data side and replaced with the prediction error signal of the coded block partially decoded. And “0” is output. Further, when the synchronizing word is detected by the synchronizing word detector 309, the switching switch 308 is switched to the inverse discrete cosine converter 303, and a prediction error signal is output. That is, when the specific code is detected by the specific code detector 307, the changeover switch 308 is switched to “0” and the synchronizing word is detected by the synchronizing word detector 309. Is detected, the switching switch 308 is switched to the inverse discrete cosine transformer 303 side. By performing such control, when a specific code is detected by the specific code detection unit 307, “0” data is output from the switching switch 308, and the data is output from the frame memory 304 before Since the image at the same position in the frame is output, the reproduced image has the same encoding block as the previous frame. This is an example of concealment processing that replaces an image in the same position as the previous frame when an error occurs in the coded block being processed. This concealment process is continued until the synchronization word is detected by the synchronization word detection unit 309.

When the synchronization word is detected by the synchronization word detector 309, a prediction error signal is output from the switch 308, and a prediction signal is output from the frame memory 304. Then, these are added by the adder 306 to obtain a reproduced image. FIG. 9 is a diagram showing an example of the operation of the image communication device according to the present embodiment. For example, encoding blocks # 1 to # 6 are made into one video packet, and are divided into three AL-SDU image packets. That is, a part of coding block # 1, # 2 and coding block # 3, the rest of coding block # 3, a part of coding block # 4 and coding block # 5, and a coding block # 5 And encoding block # 6. Each part has an error detection The output code is assigned. In addition, a synchronization word is added to the head of each transmission unit. Then, for example, if a specific code is detected in the remaining portion of the coding block # 3, the coding block in which the specific code is detected, that is, from coding block # 3 to the next synchronizing word is subjected to the concealment processing, and As described above, according to the image communication terminal device according to the present embodiment, when the specific code is detected from the image signal separated by media separation unit 104, Stops outputting the prediction error signal of the decoded coded block, outputs the image at the same position in the previous frame held in the frame memory 304, and detects this output to detect the next synchronization word Therefore, the image quality degradation when a transmission error occurs can be further suppressed as compared with the image communication terminal apparatus according to the second embodiment.

Note that, in the present embodiment, it has been described that replacement is performed with a coding block located at the same position as the previous frame. However, since the present invention is an invention for specifying a concealment processing target, a concealment method (for example, It can easily be inferred that it does not depend on the replacement method using the coded block information.

Further, although the first to third embodiments of the present invention have a hardware configuration by taking an image communication terminal device as an example, they can be realized by software. That is, the method of the present invention is programmed and stored in a writable recording medium such as a semiconductor memory, a magnetic disk, an optical disk, and a magneto-optical disk, and the stored program is processed by a CPU (central processing unit). You may try to do it. In addition, the present invention can read out the software from the recording medium and execute the software on a computer, and it goes without saying that the same operation is performed.

Further, the present invention can of course be applied to a base station device of a mobile communication system.

In addition, if the transmission apparatus includes the first to third embodiments of the present invention and has a configuration including a multiplexing unit and an encoding unit and is used on the transmission side and the reception side, bidirectional wireless image communication can be performed. Further, the present invention is particularly effective in a wireless transmission path in which an error easily occurs during transmission, but is also effective in a wired transmission path. As described above, according to the present invention, when an error is detected in an image signal in media separation processing, it is possible to easily notify an image decoding unit of a position where an error has occurred in AL-SDU image packet units. Become.

Also, when a specific code is detected in the received image signal, it is possible to display an image that could be decoded just before the error occurred, and to perform concealment processing from the time immediately after the error occurred until the next synchronization. Becomes possible. As described above, since the error position can be specified while using the variable-length code, the display of the correctly decoded portion and the portion to be concealed can be realized efficiently and with simple means. In other words, it is possible to achieve high image quality when transmission errors occur with extremely few hardware or software resources. This description is based on Japanese Patent Application No. 2000-124324, filed on April 25, 2000. This content is included here. Industrial applicability

The present invention is suitable for use in a moving image communication device such as a videophone.

Claims

The scope of the claims

1.1 After dividing an image signal for one image into a plurality of coding blocks, a video packet is generated with a plurality of coding processes as one transmission unit, and the video packet is generated as it is or in image packet units generated by dividing the video packet. A media separation method that performs error detection and replaces an erroneous image packet with a specific code that can be specified by image decoding processing.

2.1 After dividing an image signal for one image into a plurality of coding blocks, generate a video packet using a plurality of coding programs as one transmission unit, and include an image packet generated by splitting or dividing the video packet as it is. When the image packet is separated from the multiplexed signal, if there is an erroneous image packet, the image packet is discarded and inserted instead into a specific code that can be specified in the image decoding process. Separation method.

3. After dividing an image signal for one image into a plurality of coding blocks, a video packet is generated in which a plurality of coding processes are used as one transmission unit, and the video packet includes an image packet generated as it is or by dividing the video packet as it is. Separating the image packet from the multiplexed signal;

An error detection step for detecting an error in the image packet;

An image packet discarding step of discarding an image packet when an erroneous image packet is detected;

A specific code input step for inserting a specific code that can be specified in the image decoding process in place of the discarded image packet,

A media separation method, comprising:

4. The media separation according to any one of claims 1 to 3, wherein the specific code is a code that is not included in a variable-length code template including various codes for representing an image. Method.

5. After dividing the image signal for one image into a plurality of coding blocks, generate a video packet with a plurality of coding blocks as one transmission unit, and generate the video packet as it is or by dividing this video packet. From the multiplexed signal containing Separating means for separating image packets;

Error detecting means for detecting an error in the image packet;

Image packet discarding means for discarding an image packet when an erroneous image packet is detected;

Specific code insertion means for inserting a specific code that can be specified in image decoding processing in place of the discarded image packet,

A media separation device comprising:

6. The specific code inserting means inserts a code that is not in a variable length code table including various codes for representing an image, as a specific code, instead of an erroneous image packet. A media separation device as described.

7. A recording medium which stores a media separation program and is readable by a computer, wherein the media separation program divides an image signal for one image into a plurality of coding blocks, and then generates a plurality of codes. A video packet with a multiplexed block as one transmission unit, and a separation procedure for separating an image signal from a multiplexed signal including an image packet generated by directly or dividing the video packet, and an error for detecting an error in the image packet. A detection procedure, a discard procedure for discarding an erroneous image bucket when it is detected, and a specific code insertion procedure for inserting a specific code identifiable in an image decoding process in place of the discarded image bucket. A recording medium characterized by comprising:

8. A specific code that is not a variable-length code during decoding while decoding an image signal divided into a plurality of encoding blocks using a variable-length code in a variable-length code tuple. The image decoding method characterized in that, if there is, the position of the coding block immediately before the specific code is specified, and the coding block before the specified coding block has been correctly recognized.

9. A decoding step of decoding an image signal divided into a plurality of encoding blocks using a variable length code in a variable length code table;

A specific code detection step of detecting a specific code that is not a variable length code during decoding; and, if the specific code is detected during decoding, a code immediately before the detected specific code. A coding block position specifying step of specifying a position of a coding block.

After dividing an image signal for one image into a plurality of coding blocks, each of the divided coding blocks is coded by a variable length code, and a plurality of variable length coded blocks are transmitted as one. As a unit, an image signal with a synchronization word added to the beginning of each transmission unit is decoded. If a specific code that is not a variable-length code is detected during decoding, the next coded block in which the specific code is detected is used. An image decoding method characterized in that up to the synchronous word is subjected to concealment processing.

1.1.1 After dividing an image signal for one image into a plurality of coding blocks, each of the divided code blocks is coded by a variable length code, and the plurality of variable length coded blocks are converted into one transmission unit. A decoding step for decoding an image signal in which a synchronization word is added at the beginning of each transmission unit;

A specific code detection step of detecting a specific code that is not a variable length code during decoding of the image signal;

When the specific code is detected, a coding block position specifying step for specifying the position of the coding block to which the specific code is added;

When the position of the encoding block to which the specific code is given is specified, a concealment processing target position specifying step of determining a concealment processing target position based on the specified position and the next synchronization word,

An image decoding method comprising:

1 2. Decoding means for decoding an image signal divided into a plurality of encoded blocks using a variable-length code in a variable-length code tuple,

A specific code detecting means for detecting a specific code that is not a variable length code during decoding; and a code for specifying a position of an encoding block immediately before the detected specific code when the specific code is detected during decoding. Block position specifying means,

An image decoding device comprising:

13.1. After dividing an image signal for one image into a plurality of encoding programs, each of the divided encoding blocks is encoded using a variable length code, and a plurality of variable length encoded codes are encoded. Decoding means for decoding an image signal in which a synchronization word is added to the beginning of each transmission unit, using the coding block as one transmission unit;

Specific code detection means for detecting a specific code that is not a variable length code during decoding of the image signal;

When the specific code is detected, a coding block position specifying means for specifying a position of the coding block to which the specific code is added;

When the position of the encoding block to which the specific code is assigned is specified, a concealment processing target position specifying unit that determines a concealment processing target position based on the specified position and the next synchronization word,

An image decoding device comprising:

14. A recording medium storing an image decoding program and being readable by a computer, wherein the image decoding program converts an image signal divided into a plurality of encoding blocks into a variable length code in a variable length code table. A decryption procedure for performing decryption using

A specific code detection procedure for detecting a specific code that is not a variable length code during decoding, and a code for specifying a position of an encoded block immediately before the detected specific code when the specific code is detected during decoding. Block location identification procedure;

A recording medium comprising:

15. A recording medium which stores an image decoding program and is readable by a computer, wherein the image decoding program divides an image signal for one image into a plurality of encoding blocks, and then divides each of the divided codes. A decoding procedure for encoding the encoded block using a variable-length code, and using a plurality of variable-length-encoded blocks as one transmission unit to decode an image signal with a synchronization word added to the beginning of each transmission unit; A specific code detection procedure for detecting a specific code that is not a variable length code during decoding of the image signal;

When the specific code is detected, a coding block position specifying procedure for specifying a position of the coding block to which the specific code is added;

When the position of the encoding block to which the specific code is given is specified, Concealment processing target position specifying procedure for determining the concealment processing target position based on the

A recording medium comprising:

16. An image communication terminal device comprising the media separation device according to claim 5 or 6.

17. A base station device comprising the media separation device according to claim 5 or 6.

18. An image communication terminal device comprising the image decoding device according to claim 12 or claim 13.

19. A base station device comprising the image decoding device according to claim 12 or claim 13.