JPH11205152A - Data encoding device, data multiplexing device and data encoding and multiplexing device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To easily control the transmission rate of encoded data to be multiplexed, corresponding to the quality of a transmission line, by outputting the position information of respective coefficients in a variable length encoded orthogonal transformation coefficient sequence together with the variable length encoded data. SOLUTION: First of all, a VLC part 15 transforms quantized DCT coefficient data D14 to the combination of run levels. Next, the VLC part 15 performs variable length encoding processing to the run levels based on a prescribed transformation table and outputs the result to a buffer 16 as variable length code data D10A. At the same time, a position signal D15 showing the leading bits of the DC coefficient, respective DCT coefficients and EOB coefficient of the variable length code data D10A is supplied to a coefficient position output part 17 synchronously with the variable length code data D10A. Based on the position signal D15, the coefficient position output part 17 outputs a coefficient position signal D16A and a reset signal D17A to the buffer 16 synchronously with the variable length code data D10A.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

[Table of Contents] The present invention will be described in the following order.

BACKGROUND OF THE INVENTION Prior Art (FIG. 9) Problems to be Solved by the Invention Means for Solving the Problems Embodiments of the Invention (1) First Embodiment (FIGS. 1 to 4) (2) Second embodiment (FIGS. 5 to 8) (3) Other embodiments Effects of the invention

[0003]

The present invention relates to a data encoding device,
Regarding a data multiplexer and a data encoding multiplexer,
For example, the present invention is suitably applied to a data encoding device and a data multiplexing device used for digital multiplex broadcasting.

[0004]

2. Description of the Related Art Hitherto, various compression encoding methods have been proposed as methods for reducing the amount of video and audio information, and a typical one is MPEG2 (Moving Picture Experts).
There is a method called GroupPhase 2). MPEG
2. Description of the Related Art Digital broadcasting systems have been started in which video and audio broadcast data are compression-encoded using two methods and broadcast using terrestrial waves and satellite waves. In the digital broadcasting system, the coded broadcast data is compressed and coded and packetized for each predetermined block, and the resulting packet sequence is transmitted. In the digital broadcasting system, the broadcast data is packetized as described above, whereby the broadcast data of a plurality of channels is multiplexed, whereby a plurality of channels can be broadcast on one line.

FIG. 9 shows an example of an encoding device and a multiplexing device. N encoding devices 4 A to 4 N having the same configuration are connected to the multiplexing device 5. Encoding devices 4A to 4N receive broadcast data D1A to D1N from corresponding broadcast data supply means (not shown) such as a video tape recorder.
The encoding devices 4A to 4N compress and encode the corresponding broadcast data D1A to D1N according to the MPEG2 system, and encode the encoded data D in packet form.
The signals are supplied to the multiplexer 5 as 4A to 4N.

[0006] The multiplexing device 5 includes encoded data D4A to D4D.
4N is multiplexed and supplied to a transmitting device (not shown) as multiplexed transmission data D5.

[0007]

In such a digital broadcasting system, when the quality of the transmission path is degraded due to a typhoon, heavy rain, or the like, the transmission side increases the transmission output of the broadcast signal to increase the transmission power on the reception side. The signal-to-noise ratio (S / N ratio) is prevented from lowering. However, when the quality of the transmission path is extremely deteriorated, there is a case where a stable reception state cannot be ensured even if the transmission output is increased. In this case, the viewer may not be able to view the broadcast.

As one method for solving such a problem, a method for enhancing the error correction capability can be considered. That is, the error correction capability is enhanced by increasing the information amount of the error correction code (parity) added to the transmission data, so that even if the quality of the transmission path is deteriorated, the error correction code on the receiving side is sufficient. , The error correction process can be performed, and the viewer can view the broadcast.

However, if the transmission rate of the transmission path has an upper limit in such a digital broadcasting system, the amount of information that can be transmitted, that is, the amount of coded data, is reduced due to the enhancement of the error correction capability. This means that the quality of the transmitted image is degraded. Therefore, it is not practical to always use a strong error correction capability. For this reason, it is necessary to adaptively change the error correction capability according to the quality of the transmission path, and to always perform optimal error correction processing.

[0010] The present invention has been made in view of the above points, and a data coding apparatus and a data multiplexing apparatus capable of easily adjusting the transmission rate of coded data to be multiplexed according to the quality of a transmission path. A device and a data encoding and multiplexing device are proposed.

[0011]

According to the present invention, in order to solve the above-mentioned problems, the data encoding means converts the position information of each coefficient of the variable-length-coded orthogonal transform coefficient sequence into a variable-length coded data. And the data multiplexing means determines the coefficient position of the variable-length coded data output from the data coding means based on the position information of each coefficient output together with the variable-length coded data. Thus, the multiplexing means can easily detect the coefficient position.

Therefore, when the multiplexing means reduces the data amount of the variable-length coded data in accordance with a change in transmission quality or a change in the priority assigned to each channel, the multiplexing means inputs the data as the variable-length coded data. It is possible to easily discard the predetermined order term of the orthogonal transform coefficient sequence to be performed.

[0013]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below in detail with reference to the drawings.

(1) First Embodiment In FIG. 1, reference numeral 50 denotes a multiplexing apparatus, to which N encoding apparatuses 10A to 10N having the same configuration are connected.
Encoding devices 10A to 10N are supplied with image data D1A to D1N as broadcast data from broadcast data supply means (not shown) such as a corresponding video tape recorder.

Here, the encoder 10A inputs the image data D1A to the preprocessor 11 as shown in FIG. The preprocessing unit 11 specifies which of the three image types of I-picture, P-picture or B-picture is to be processed for each frame image of the sequentially input image data D1A, and then specifies the image type of the frame image. The frame images are rearranged in the encoding order according to the image type, and the frame images are further divided into macroblocks composed of a luminance signal of 16 pixels × 16 lines and a color difference signal corresponding to the luminance signal. This is supplied to the arithmetic circuit 12 and the motion vector detector 23 as macroblock data D2.

The motion vector detecting section 23 calculates the motion vector of each macro block of the macro block data D2 based on the macro block data D2 and the reference image data D20 stored in the frame memory 21. The data is sent to the motion compensation unit 22 as D23.

The arithmetic circuit 12 performs an intra mode, a forward prediction mode, and a backward prediction mode on the macroblock data D2 supplied from the preprocessing unit 11, based on the image type of each macroblock of the macroblock data D2. Alternatively, motion compensation is performed in one of the bidirectional prediction modes. Here, the intra mode is a method in which a frame image to be encoded is directly used as transmission data, and the forward prediction mode is a method in which a prediction residual between the frame image to be encoded and a past reference image is transmitted data. It is a method. In addition, the backward prediction mode is a method in which a prediction residual between a frame image to be encoded and a future reference image is used as transmission data, and the bidirectional prediction mode is
This is a method in which a prediction residual between a frame image to be encoded and an average value of two prediction images of a past reference image and a future reference image is used as transmission data.

First, the case where the macroblock data D2 is an I-picture will be described. In this case, the macro block data D2 is processed in the intra mode. That is, the arithmetic circuit 12 converts the macro block of the macro block data D2 into the DCT
(Discrete Cosine Transform). The DCT unit 13 performs DCT conversion processing on the operation data D3 to convert the operation data D3 into DCT coefficients, and sends the result to the quantization unit 14 as DCT coefficient data D13. The quantization unit 14 performs a quantization process on the DCT coefficient data D13, and outputs VD as the quantized DCT coefficient data D14.
It is sent to the LC unit 15 and the inverse quantization unit 18. At this time, the quantization unit 14 controls the generated code amount by adjusting the quantization step size in the quantization process according to the quantization control value D24 supplied from the control unit 24.

The quantized DCT sent to the inverse quantization unit 18
The coefficient data D14 undergoes an inverse quantization process, and is sent to the inverse DCT unit 19 as DCT coefficient data D18. Then, the DCT coefficient data D18 undergoes inverse DCT processing in the inverse DCT section 19, and becomes the arithmetic circuit 2
0 and is stored in the frame memory 21 as reference image data D20.

Next, the case where the macroblock data D2 is a P-picture will be described. In this case, the arithmetic circuit 12 performs a motion compensation process on the macroblock data D2 in one of the intra mode and the forward prediction mode.

When the prediction mode is the intra mode, the arithmetic circuit 12 sends the macroblock of the macroblock data D2 as it is to the DCT unit 13 as the arithmetic signal D3, as in the case of the I-picture.

On the other hand, when the prediction mode is the forward prediction mode, the arithmetic circuit 12 outputs the macroblock data D2.
Is subtracted using the forward prediction image data D22F supplied from the motion compensation unit 22.

The forward prediction image data D22F is the reference image data D20 stored in the frame memory 21.
Is calculated by performing motion compensation according to the motion vector data D23. That is, in the forward prediction mode, the motion compensator 22 shifts the read address of the frame memory 21 in accordance with the motion vector data D23 to read the reference image data D20, and uses the read reference image data D20 as the forward prediction image data D22F. 20. The operation circuit 12 subtracts the forward prediction image data D22F from the macroblock data D2 to obtain difference data as a prediction residual, and obtains the DCT unit 13 as operation data D3.
To send to.

Further, the forward prediction image data D22F is supplied from the motion compensator 22 to the arithmetic circuit 20, and the arithmetic circuit 20 adds the forward prediction image data D22F to the arithmetic data D19 to obtain the reference image data. D2
0 (P picture) is locally reproduced and stored in the frame memory 21.

Next, the case where the macroblock data D2 of B picture is supplied from the preprocessing unit 11 to the arithmetic circuit 12 will be described. In this case, the arithmetic circuit 12 performs any one of the intra mode, the forward prediction mode, the backward prediction mode, and the bidirectional prediction mode on the macroblock data D2.

When the prediction mode is the intra mode or the forward mode, the macroblock data D2 is subjected to the same processing as in the case of the P picture described above. However, since the B picture is not used as another predicted reference image, the reference image data D20 in this case is not stored in the frame memory 21.

On the other hand, when the prediction mode is the backward prediction mode, the arithmetic circuit 12 outputs the macroblock data D2.
Is subtracted using the backward prediction image data D22B supplied from the motion compensation unit 22.

The backward predicted image data D22B is the reference image data D20 stored in the frame memory 21.
Is calculated by performing motion compensation according to the motion vector data D23. That is, in the backward prediction mode, the motion compensation unit 22 reads the reference image data D20 by shifting the read address of the frame memory 21 in accordance with the motion vector data D23, and uses the read reference image data D20 as the backward prediction image data D22B. 20. The arithmetic circuit 12 subtracts the backward prediction image data D22B from the macroblock data D2 to obtain difference data as a prediction residual, and obtains the DCT unit 13 as operation data D3.
To send to.

Further, the backward prediction image data D22B is supplied from the motion compensator 22 to the arithmetic circuit 20, and the arithmetic circuit 20 adds the backward prediction image data D22B to the arithmetic data D19 to obtain the reference image data. D2
Although 0 (B-picture) is locally reproduced, the B-picture is not used as another predicted reference image, so the reference image data D20 in this case is not stored in the frame memory 21.

When the prediction mode is the bidirectional mode, the arithmetic circuit 12 calculates the motion compensator 2 from the macroblock data D2.
2 is subtracted from the average value of the forward prediction image data D22F and the backward prediction image data D22B to obtain difference data as a prediction residual, and DCT is used as the calculation data D3.
To the unit 13.

Further, the forward prediction image data D22F and the backward prediction image data D22B are supplied from the motion compensator 22 to the operation circuit 20, and the operation circuit 20 adds the forward prediction image data D22F and the The reference image data D20 (B-picture) is generated by adding the average value of the backward prediction image data D22B. However, since the B-picture is not used as another prediction reference image, the reference image data D20 in this case is a frame. It is not stored in the memory 21.

Thus, the image data D1A input to the encoding device 10A undergoes motion compensation prediction processing, DCT processing and quantization processing, and is supplied to the VLC section 15 as quantized DCT coefficient data D14.

The VLC unit 15 performs a variable length encoding process on the quantized DCT coefficient data D14 based on a predetermined conversion table. Here, the leading value of the quantized DCT coefficient data D14 is a DC coefficient indicating the difference between the DC components of the image and the image immediately before the image, and the values after the DC coefficient are DCT coefficients. Note that the DCT coefficient is obtained by zigzag scanning a block of 8 pixels × 8 lines obtained by dividing the macro block into four.

First, the VLC unit 15 converts the quantized DCT coefficient data D14 shown in FIG. 3A into a continuous "0" (referred to as "Run") as shown in FIG. It is converted to a combination of coded values (this is called Level). This combination is called a run level (Run, Level), and the order starting from 1 is assigned to the DCT coefficient of each run level. Then, a portion where all the remaining quantization values of the block are "0" is defined as an EOB (End Of Block). The maximum order of the DCT coefficients in each block depends on the complexity of the image of the block.

Here, the DCT coefficient having a low order indicates the signal level of an area where the change is flat in the block, and the DCT coefficient having a high order indicates the signal level of an area where the contour or the like changes drastically. Therefore, DCT of high order
The coefficient has less influence on the image quality than the DCT coefficient having a lower order. Therefore, in this embodiment, D
By reducing the CT coefficients from DCT coefficients of higher order, the amount of information is reduced, that is, the output rate is adjusted, while minimizing the deterioration of the image quality.

Next, the VLC unit 15 performs a variable length coding process on the run level shown in FIG. 3B based on a predetermined conversion table, and buffers it as variable length code data D10A shown in FIG. 3C. 16 and a position signal D15 (FIG. 2) indicating the DC coefficient, each DCT coefficient and each leading bit of EOB of the variable length code data D10A in synchronization with the variable length code data D10A. (FIG. 2).

Based on the position signal D15, the coefficient position output section 17 buffers the coefficient position signal D16A shown in FIG. 3D and the reset signal D17A shown in FIG. 3E in synchronization with the variable length code data D10A. 16 is output. That is, the coefficient position signal D16A indicates "1" at the bit corresponding to the head of the block of each DCT coefficient, and indicates "0" at the other bits. And the reset signal D1
7A indicates "1" in the bit corresponding to the block of the DC coefficient and the EOB, and "0" in the other bits.
Is shown. Variable length code data D10A, coefficient position signal D1
6A and the reset signal D17A are temporarily stored in the buffer 16.

The control unit 24 (FIG. 2) constantly monitors the accumulation state of the variable length code data D10A in the buffer 16, and obtains such accumulation state as the occupation amount information D11. Then, the control unit 24 generates a quantization control value D24 based on the occupation ratio information D11, sends the generated value to the quantization unit 14, and adjusts the quantization step size in the quantization process to adjust the quantization step value in accordance with the data occupation amount of the buffer. Generate the amount of data.

In FIG. 2, variable length code data D10
A, coefficient position signal D16A and reset signal D17A
Are synchronously read from the buffer 16 and sent to the multiplexer 50 shown in FIG. The multiplexing device 50 includes N [number] information amount reducing units 30A to 30N having the same configuration corresponding to the encoding devices 10A to 10N, and each of the information amount reducing units 30A to 30N corresponds to the corresponding encoding device. 10A
-10N and the multiplexing unit 40.

The variable length code data D10A, coefficient position signal D16A, and reset signal D17A output from the first encoding device 10A are synchronously supplied to the information amount reducing unit 30A. In the information amount reducing unit 30A, the coefficient position signal D16A and the reset signal D16A
1 is input.

The coefficient counter 31 outputs a coefficient position signal D16
A count operation is performed based on A and the reset signal D17A. That is, as shown in FIG.
When the bit "1" arrives for the coefficient position signal D16A, "1" is added to the count value, and the reset signal D1
When bit "1" arrives for 7A, the count is reset and the count value is set to "0". The count value is supplied to the coefficient discard control unit 32 as the coefficient count value D31. Here, the coefficient count value D31 is output in synchronization with the input of the coefficient position signal D16A and the reset signal D17A.
31 is supplied to the coefficient discarding control unit 32 in synchronization with the coefficient position signal D16A.

Here, the transmission line information D41 and the priority information D42 are supplied to the coefficient discarding control unit 32. The transmission path information D41 is transmission quality information of a transmission path in the digital broadcasting system, and is supplied from an apparatus for monitoring a signal transmission state in the digital broadcasting system by an operation of an operator or the like. The priority information D42 is information indicating the priority of each image data D1A to D1N, that is, each program in the digital broadcasting system, and is supplied from a device for managing the broadcasting contents of the digital broadcasting system by an operation of an operator or the like. Is done.

The coefficient discard control unit 32 transmits the transmission path information D41
Based on the priority information D42 and the priority information D42, the amount of the error correction signal added to the transmission signal (multiplexed data D40) is estimated, and a coefficient discard reference value is set according to the estimated amount. Here, the coefficient discard reference value is a value indicating the order of the DCT coefficient to be transmitted for the DCT coefficient of the variable-length coded data. For example, when the coefficient discard reference value is “2”, the first and second order DCT coefficients are transmitted, and the remaining third and subsequent DCT coefficients are discarded. The DC coefficient and EOB are always transmitted without being discarded.

For example, the coefficient discarding control unit 32 sets the coefficient discarding reference value high when the transmission quality of the transmission path is relatively good, and sets the coefficient discarding reference value low when the transmission quality of the transmission path is relatively poor. I do. When the priority is high, the coefficient discard reference value is set low, and when the priority is low, the coefficient discard reference value is set high.

When the coefficient count D31 exceeds the coefficient discard reference value in the coefficient discarding control unit 32, the coefficient discarding control unit 32 outputs the discard instruction signal D32 to the coefficient discarding unit 33.
Output to FIG. 4 shows the relationship between the variable length code data D10A and the discard command signal D32. In this case, the coefficient discard reference value is set to “2”, and when the order of the DCT coefficient is 3 or more in the variable-length code data D10A, a discard command signal D32 is output.

The coefficient discarding unit 33 discards the bit string of the variable length code data D10A in response to the discard command signal D32. The discarded bit string is a DCT coefficient having an order greater than the coefficient discarding reference value. The coefficient discarding unit 33 buffers the DC coefficient, the EOB, and the DCT coefficient having an order less than the coefficient discarding reference value as the encoded data D33A.
4 and the coded data D33A is temporarily stored in the buffer. Such discard processing is performed by the transmission path information D
41, and dynamically changes according to the priority information D42, and there is no delay due to the discard processing.

Although the processing of the image data D1A supplied to the encoding device 10A has been described above, the information amount reduction units 30B to 30N are also applied to the other image data D1B to D1N.
Receive the same processing.

The coded data D33A stored in the buffer 34 of the information amount reducing unit 30A is read out as coded data D30A in response to a read command from the multiplexing unit 40. Information amount reduction unit 30
Encoded data D30B to D3 transmitted from B to 30N
The signal is multiplexed with 0N, supplied to a transmission device (not shown) as multiplexed data D40, transmitted with an error correction code according to the transmission quality of the transmission path.

In the above configuration, the encoding devices 10A to 10A
10N are variable-length code data D10A to D10N, respectively.
And outputs the coefficient position information (coefficient position signals D16A to D16N) indicating the coefficient position for each order of the DCT coefficient sequence output to the multiplexer 50 together with the variable length code data D10A to D10N.

As a result, the information amount reduction units 30A to 30N of the multiplexing device 50
A-10N, variable-length code data D10A-
For each of the encoding devices 1 in synchronization with D10N,
Coefficient position signals D16A to D supplied from 0A to 10N
16N, each variable-length code data D10A to D10N
Of each order of the DCT coefficient of each variable-length code data D1
A state can be determined at the same time when 0A to D10N are input.

Therefore, the information amount reduction units 30A to 30N
When the transmission quality and / or priority of the transmission path changes, the variable length code data D10A to D10D are immediately changed according to these changes.
A 10N DCT coefficient sequence can be cut at a predetermined order position. As a result, each of the variable-length code data D10A to D10D
In the 10N, the multiplexing device 50 performs a process of increasing or decreasing the data amount substantially in real time according to a change in the transmission quality and / or the priority of the transmission path.

The variable-length code data D10A to D10N subjected to the data amount increase / decrease processing as described above are output from the multiplexer 50 as multiplexed data D40 after being multiplexed, and provided at the subsequent stage of the multiplexer 50. The error correction code adding section (not shown) adaptively adds an error correction code (parity) having a length corresponding to the transmission quality and priority of the transmission path.

As described above, even if the transmission quality and / or the priority of the transmission path changes, the data amount control is performed inside the multiplexer 50 without controlling the data amount generated in the encoders 10A to 10N. Can be done immediately.

Therefore, when the transmission quality of the transmission line is deteriorated, a large number of DCT coefficients are immediately reduced by lowering the coefficient discarding reference value in the information amount reduction units 30A to 30N, and the variable length is accordingly reduced. The data length of the error correction code added to the code data can be increased. As a result, the error correction capability of the variable-length code data increases substantially in real time according to the state of the transmission path. In this case, for the variable-length code data to which the high priority is assigned, the information amount reducing unit 3
The coefficient discard reference values set in 0A to 30N are particularly low, and the DCT coefficients to be reduced increase. Accordingly, the data length of the error correction code added to the variable-length code data can be increased by this amount, and the error correction capability of the variable-length code data becomes higher than that of the variable-length code data having a lower priority.

Thus, the interruption of the broadcast can be avoided by switching the error correction capability of each of the variable-length code data D10A to D10N multiplexed on the multiplexed data D40 in real time according to the transmission quality of the transmission line, thereby avoiding the interruption of the broadcast. Even if the transmission quality of the transmission path is extremely deteriorated by increasing the error correction capability particularly for the variable-length code data with high priority, the broadcast of the channel with high priority is reliably received on the receiving side.

According to the above configuration, the encoding devices 10A to 10A
10N, in synchronization with the quantized DCT coefficient data D14,
Coefficient position signals D16A to D16 indicating the positions of DCT coefficients
A coefficient position output unit 17 for outputting N and reset signals D17A to D17N is provided to continuously output coefficient position information to the multiplexer 50 in real time,
A coefficient discarding unit 33 for discarding DCT coefficients based on the coefficient position signals D16A to D16N, the reset signals D17A to D17N, the transmission line information D1, and the priority information D2. By discarding the DCT coefficients of the coded DCT coefficient data D14 from the higher-order DCT coefficients, even when the transmission state deteriorates, the information amount can be immediately reduced in the multiplexing device 50. , The error correction capability can be switched in real time according to the change of the error correction capability. Therefore, interruption of broadcasting can be prevented.

The multiplexing device 50 includes the encoding device 10
Coefficient position signals D16 output from A to 10N respectively
Variable length code data D10A to D10D based on A to D16N
By being able to detect the order position of the 10N DCT coefficient, the configuration of the multiplexing device 50 can be simplified compared to a method of temporarily decoding the variable length code data D10A to D10N to detect the order position of the DCT coefficient. it can.

(2) Second Embodiment At present, B-ISDN (Broadband-Integrated Services)
Digital broadcasting systems using Digital Network, a broadband integrated digital communication service network) are being considered. In B-ISDN, ATM (Asynchronous Transfer
r Mode, asynchronous transfer mode). In the ATM method, data to be transferred is 48 [Byt]
e] length, and control information such as a transmission path called a 5 [byte] header is added to this to generate a 53 [byte] ATM cell, and the ATM cell is transferred to a plurality of ATM exchanges. The data is transferred by asynchronously transferring the data on the connected ATM network. The ATM system is designed to efficiently transfer data at various transfer rates by increasing or decreasing the amount of generated ATM cells according to the transfer rate of the data to be transferred.

In FIG. 5, in which parts corresponding to those in FIG. 1 are assigned the same reference numerals, reference numeral 90 denotes an ATM network, and a plurality of ATM switches 91 are interconnected to form a network. A plurality of ATM terminals 92 are connected to each ATM exchange, and a user of the ATM network connects to the ATM network 90 via the ATM terminal 92.

In FIG. 5, reference numeral 80 denotes a multiplexer.
N encoding devices 10A to 10N having the same configuration are connected. Encoding devices 10A to 10N are supplied with image data D1A to D1N from corresponding broadcast data supply means (not shown) such as a video tape recorder. The configurations and operations of the encoding devices 10A to 10N are the same as those in FIG. 2 of the first embodiment.

The image data D1A input to the encoding device 10A undergoes a motion compensation prediction process, a DCT spreading process, a quantization process, and a variable length encoding process in the encoding device 10A. The signal is sent to the multiplexer 80 in synchronization with the coefficient position signal D16A and the reset signal D17A.

In FIG. 6, a multiplexing device 80 includes N [number] data partitioning units 60A to 60N corresponding to the encoding devices 10A to 10N.
0, the variable-length code data D10A, the coefficient position signal D16A, and the reset signal D17A are supplied to the data partition unit 60A.

In the data partitioning section 60A,
The coefficient position signal D16A and the reset signal D17A are input to a coefficient counter 61.

The coefficient counter 61 calculates the coefficient position signal D16
A count operation similar to that of the first embodiment is performed based on A and the reset signal D17A, and the count value is supplied to the coefficient separation control unit 62 as a coefficient count value D61.

Here, the coefficient separation control section 62 is supplied with congestion information D81 and priority information D82. The congestion information D81 is information indicating the congestion state (congestion state of the line, etc.) of the ATM network, and the priority information D2 is the image data D1A.
D1N, that is, information indicating the priority of each program in the digital broadcasting system. Coefficient separation control unit 62
Sets a priority reference value as a reference for the data partitioning process based on the congestion information D81 and the priority information D82.

The data partitioning process is an MP
This is a process of dividing each DCT coefficient of the encoded data into two groups according to the importance defined by the EG2 method. A group having a high importance is called a partition 0,
The group of low importance is called Partition 1. Such separation processing is performed based on a value called a priority reference value (Priority Breakpoint). When the priority reference value is “64”, the DCT coefficient 1 is set to partition 0, and the DCT coefficient 2 and thereafter are set to partition 1. If the priority reference value is "6
In the case of "5", the DCT coefficient 1 and the DCT coefficient 2 are set to partition 0, and the DCT coefficient 3 and thereafter are set to partition 1. Thereafter, each time the priority reference value increases by one, the order of the DCT coefficient included in the partition 0 increases by one. Thus, the order of the DCT coefficients included in the partition 0 is determined according to the priority reference value. The DC coefficient is included in partition 0, and the EOB is included in partition 1.

FIG. 7 shows an example of data partitioning, in which the encoded data shown in FIG. 7A is separated into a partition 0 shown in FIG. 7B and a partition 1 shown in FIG. 7C. doing. In this case, the priority reference value is set to 64, the partition 0 includes the DC coefficient and the DCT coefficient 1, and the partition 1 includes the DCT coefficient 2 and the subsequent and the EOB. Although the video can be reproduced by using only the partition 0, the image quality is deteriorated as compared with the video reproduced by using the partition 0 and the partition 1. A video cannot be reproduced only by the partition 1.

In the coefficient separation control section 62, when the coefficient count value D61 exceeds the order corresponding to the priority reference value,
The coefficient separation control section 62 supplies a separation command signal D62 to the coefficient separation section 64. The coefficient separating section 64 normally selects the buffer (0) 66 side, and selects the buffer (1) 67 side only when the separation command signal D62 is supplied. Thus, the variable-length code data D10A is separated into two partitions, partition 0 and partition 1, according to the congestion information D81 and the priority information D82, and the encoded data D64 (partition 0) and the encoded data D
65 (partition 1) is temporarily stored in the corresponding buffer (0) 66 and buffer (1) 67, respectively. Such separation processing dynamically fluctuates according to the congestion information D81 and the priority information D82, and no delay occurs due to such separation processing.

Although the processing of the image data D1A supplied to the encoding device 10A has been described above, the same processing is performed on the other image data D1B to D1N.

The coded data D64 and the coded data D65 stored in the buffer (0) 66 and the buffer (1) 67 of the data partitioning section 60A are multiplexed by the multiplexing section 70.
Are read out as encoded data D66A (partition 0) and encoded data D67A (partition 1) in response to the read command from the multiplexing unit 70, and the other data partitioning units 60B to 60B in the multiplexing unit 70.
N is multiplexed with the coded data D66B to D66N and the coded data D67A to D67N sent from the N. The multiplexed data D70 (partition 0) and the multiplexed data D
The data is sent to the ATM interface 93 shown in FIG. 5 as 71 (partition 1).

At this time, in the multiplexing unit 70, the multiplexed data D70 and the multiplexed data D71 are processed according to the priority. That is, for multiplexed data D70 composed of partition 0 with a high priority,
The multiplexed data D70 can be reliably obtained by enhancing the error correction capability as compared with the multiplexed data D71 composed of the partition 1 having a low priority, or by selecting a transmission path having higher transmission quality. Even when the transmission is performed and the transmission state is deteriorated, sudden image interruption is prevented, and a gradual deterioration in image quality, that is, a so-called graceful degradation characteristic can be realized.

The multiplexed data D70 and the multiplexed data D7
1 is A in the ATM interface 93 shown in FIG.
After receiving the TM cell processing, the ATM cell D93
It is sent to the ATM network 90 via the M terminal 92 and
Via the exchange 91, it reaches the target ATM terminal 92.

At this time, by associating the CLP (Cell Loss Priority) bit described in the header of the ATM cell with the partition, even if congestion occurs, the ATM cell of the partition 0 can be reliably obtained. Can be transmitted to Here, the CLP bit is priority information of a cell used for controlling the ATM cell flow in the ATM network, which is defined by the ATM system. When congestion occurs in the ATM network, the ALP having a lower priority in the ATM exchange is used.
Congestion is eliminated by discarding TM cells. That is, as shown in FIG. 8, the CLP bit is set to "0" indicating that the priority is high for the ATM cell of the partition 0, and the CLP bit is indicated to indicate that the priority is low for the ATM cell of the partition 1. By setting "1", when congestion occurs, the ATM cells of the partition 1 having the CLP bit "1" are preferentially discarded in the ATM exchange, so that sudden video cutoff can be prevented even in a congested state.

The ATM cell D93 that has reached the target ATM terminal 92 is decoded by the ATM interface 93 and supplied to the video receiving device 85 as encoded data D94. The encoded data D94 is subjected to a decoding process by the MPEG2 system in the video receiving device 85, and is supplied as video data to a television receiver (not shown).

In the above configuration, the encoding devices 10A to 10A
10N are variable-length code data D10A to D10N, respectively.
And outputs the coefficient position information (coefficient position signals D16A to D16N) indicating the coefficient position for each order of the DCT coefficient sequence output to the multiplexer 80 together with the variable length code data D10A to D10N.

The data partitioning units 60A to 60N of the multiplexing device 80 correspond to the corresponding encoding devices 10A.
To variable length code data D10A to D10D supplied from
10N, each encoding device 10
Coefficient position signals D16A to D1 supplied from A to 10N
Using 6N, the order positions of the DCT coefficients of the variable length code data D10A to D10N can be determined.

Accordingly, the data partitioning units 60A to 60A
60N is input variable-length code data D10A to D10D.
Once the variable length decoding of 10N is performed and the order of each DCT coefficient is determined, complicated processing is not required, and as a result, data partitioning processing can be easily performed.

According to the above configuration, the encoding devices 10A to 10A
By supplying the coefficient position signals D16A to D16N to the multiplexer 80 in addition to the variable length code data D10A to D10N from 10N, the data partitioning process in the multiplexer 80 can be easily performed.

(3) Other Embodiments In the above-described first embodiment, two types of signals, the coefficient position signal D16A and the reset signal D17A, are transmitted. Not limited,
A reset instruction may be superimposed on the coefficient position signal.

An example of the superposition method will be described below. MPEG
Since the bit length of the variable-length code after the variable-length coding is specified to be 2 bits or more in the two-system, "1" indicating the head of the DCT coefficient is used in the coefficient position signal D16A shown in FIG. Are not continuous. Therefore, the reset instruction can be superimposed on the coefficient position signal by setting the reset instruction using the bit "11" in the coefficient position signal. That is, FIG. 3 (G) shows the coefficient position / reset signal D26A. The bit corresponding to the first two bits of each DCT coefficient block is set to "10", and the EOB is set.
The bit corresponding to the first two bits of the block is "11".
And the other bits are set to 0. In the coefficient counter 31 (FIG. 1), the coefficient position / reset signal D2
By incrementing the counter by "1" when the bit string "10" is input for 6A and by resetting the counter when the bit string "11" is input, the information amount reducing unit 30A can determine the coefficient position Reset signal D2
The coefficient discarding process can be performed with reference to only 6A.

In the first embodiment described above,
The coefficient discarding control unit 32 discards the DCT coefficient by controlling the coefficient discarding unit 33 based on the transmission path information D41 and the priority information D42. However, the present invention is not limited to this, and other various information may be used. May be used to control the coefficient discarding unit 33.

In the above-described second embodiment,
The coefficient separation control unit 62 performs the data partitioning process by controlling the coefficient separation unit 64 based on the congestion information D81 and the priority information D82. However, the present invention is not limited to this, and other various information may be used. May be used to control the coefficient separation unit 33.

In the first and second embodiments described above, the case where image data is transmitted has been described. However, the present invention is not limited to this, and the case where various other data such as audio data is transmitted is described. Can be applied to

[0084]

As described above, according to the present invention, the data encoding means outputs the position information of each coefficient of the variable-length-coded orthogonal transform coefficient sequence together with the variable-length-coded data, In the data multiplexing means, by detecting the coefficient position of the variable length coded data output from the data coding means based on the position information of each coefficient output together with the variable length coded data, ,
The multiplexing means can easily detect the coefficient position.

Thus, the multiplexing means reduces the data amount of the variable-length coded data according to the change in the transmission quality of the transmission path and the change in the priority assigned to each channel. It is possible to easily discard a predetermined order term of the orthogonal transform coefficient sequence input as the converted data.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a configuration of an encoding device and a multiplexing device according to the present invention.

FIG. 2 is a block diagram showing a configuration of an encoding device according to the present invention.

FIG. 3 is a schematic diagram illustrating a relationship between a DCT coefficient and a coefficient position signal.

FIG. 4 is a schematic diagram illustrating a relationship between a DCT coefficient and a discard instruction.

FIG. 5 is a schematic diagram illustrating a configuration of an ATM network.

FIG. 6 is a block diagram showing a configuration of an encoding device and a multiplexing device according to a second embodiment.

FIG. 7 is a schematic diagram used for describing a data partitioning process.

FIG. 8 is a schematic diagram showing cell discarding in an ATM exchange;

FIG. 9 is a block diagram showing a conventional encoding device and multiplexing device.

[Explanation of symbols]

10A to 10N encoding device, 17 coefficient position output unit, 30A to 30N information amount reducing unit, 31 coefficient counter, 32 coefficient discard control unit, 33 coefficient discard unit, 40 ... multiplexing unit, 50 ... multiplexing device, 60A ~
60N: Data partitioning unit, 70: Multiplexing unit, 80: Multiplexing device.

Claims (5)

[Claims] 1. A data encoding apparatus for generating a coefficient sequence of an orthogonal transform code by orthogonally transform-encoding input data, and performing variable-length encoding on the coefficient sequence and outputting the coefficient sequence. A data encoding apparatus comprising: coefficient position information generating means for outputting information together with the variable length encoded data. 2. A method according to claim 1, wherein said coefficient position information generating means outputs a signal representing a head of said variable length code corresponding to each of said coefficients as said position information in synchronization with said variable length coded data. 2. The data encoding device according to claim 1, wherein: 3. The coefficient position information generating means outputs a signal indicating an end point of the block of the coefficient sequence orthogonally coded for each predetermined block in synchronization with the variable length coded data. The data encoding device according to claim 1, wherein the data encoding is performed. 4. The variable-length-encoded output data of a data encoding apparatus for generating a coefficient sequence of an orthogonal transform code by orthogonally transform-coding input data, and performing variable-length encoding on the coefficient sequence for output. In the data multiplexing device that inputs and multiplexes and outputs a plurality of data, the coefficient position of the variable-length coded output data output from the data coding device is output together with the output data from the data coding device. A data multiplexing device comprising a coefficient position detecting means for detecting based on output coefficient position information. 5. An orthogonal transform coding of input data to generate a coefficient sequence of an orthogonal transform code, and a variable length coding of the coefficient sequence to output the data sequence and output from a plurality of the data coding devices. A data encoding and multiplexing apparatus having multiplexing means for multiplexing and outputting the output data to be output, wherein the data encoding means outputs position information of each coefficient of the coefficient sequence together with the variable-length encoded data. The data multiplexing means outputs the coefficient position of the variable length coded data output from the data encoding means to the variable length coded data from the data encoding means. Data multiplexing apparatus comprising coefficient position detecting means for detecting based on position information of each of the coefficients output together with the generated data.