KR101500300B1 - Selective Low-Power Video Codec with Interaction Between Encoder and Decoder, and an Encoding/Decoding Method Thereof - Google Patents

Abstract

The present invention relates to a pixel domain Wyner-Ziv video codec which allows an encoder to selectively perform encoding based on cost information calculated in auxiliary information created by a decoder in an environment capable of interaction between the video encoder and the video decoder, and to a method for encoding and decoding using the same. The method according the present invention is devised to improve the limited performance of an existing pixel domain Wyner-Ziv video encoding technique, and adopts a method which classifies a Wyner-Ziv frame into AWZ and PWZ, and selectively performs encoding. As a result, a high quality image service can be achieved at a low bit rate by improving encoding efficiency, and the method can be applied to an environment capable of interaction between an encoder and a decoder which are located nearby. A method which can encode a Wyner-Ziv frame by classifying the Wyner-Ziv frame into I/AWZ/PWZ is introduced. For a method which selectively selects a block when a PWZ frame is encoded, a method which selects block candidates for a current frame by applying bit-error characteristics obtained in a recovery process in AWZ to each block is introduced. Also, the method according to the present invention comprises the steps of: measuring time matching costs; measuring costs by measuring the uniformity of a motion vector; sorting the blocks by cost value size; and sorting the blocks by combined size after sorting.

Description

TECHNICAL FIELD [0001] The present invention relates to a selective low-power video codec capable of interacting between an encoder and a decoder, and an encoding / decoding method using the same,

The present invention relates to an optional low-power video codec device capable of interacting between an encoder and a decoder, and a coding and decoding method using the same. More particularly, And a coding and decoding method using the Wiener-Jive video codec apparatus.

Digital video compression technology, which is widely used until now, requires a large amount of computation and complexity in compression encoding. For this reason, it is very limited to applications such as wireless video devices, low-power video devices, and wireless sensor network sensors in which resources such as memory and CPU computation are limited and power is very limited. Distributed video coding (DVC) is being studied as an application to such an environment.

In order to maintain a low bit rate while maintaining a good image quality, a distributed video coding technique channel-encodes information generated by simply quantizing an original image or discarding an appropriate amount of bit planes at a transmitting side, and transmits the generated parity bits. On the receiving side, And a parity bit is requested only as much as necessary through the feedback channel to obtain an appropriate performance efficiency.

That is, in the video decoder, auxiliary information, which is a predicted value image of the original image, is generated, and noise generated in the auxiliary information and the original is corrected through received parity information.

However, since only a simple operation is implemented in the encoder, the amount of computation is very small, but it is limited to greatly improve the coding efficiency.

Korean Patent Laid-Open No. 10-2010-0093703 (titled " Distributed Moving Picture Protector and Decoder < RTI ID = 0.0 > Korean Patent No. 10-1071725 (name: distributed video codec using motion information feedback) Korean Patent No. 10-1071724 (entitled: Pixel-area Distributed Video Codec Device, Coding and Decoding Method Using the Same)

SUMMARY OF THE INVENTION The present invention has been conceived to solve the problems described above, and it is an object of the present invention to provide a video encoding apparatus and a video encoding apparatus capable of selectively encoding an encoder based on cost information calculated from auxiliary information generated by a decoder in an environment in which a video encoder and a decoder can interact And a pixel region winner-jive video codec apparatus.

It is another object of the present invention to provide a pixel region Winner-Zig video encoding and decoding method for obtaining excellent performance in an environment in which an encoder and a decoder can interact with a distributed video codec applicable to an existing low-power environment.

The technical problems that the present invention devised in this way are to be accomplished are not limited to the above-mentioned technical problems.

A method of encoding a pixel region Winner-Zib video codec according to an embodiment of the present invention is a method of encoding a pixel region Winner-Ziv video codec, comprising the steps of: (AWZ) in which a video frame to be transmitted is coded in accordance with an intra picture coding scheme (I), an entire block is regarded as a Wiener-jive frame according to the received channel coding scheme information, And a method of channel encoding by considering a Wiener-Jive frame as a channel coding method (PWZ).

In this case, in the encoding method of the pixel region Winner-Jive video codec, since the encoder encodes only a part of the blocks as a Wiener-Jive frame and performs channel encoding, the channel encoding method information includes a winner- The block candidates for the current Wiener-Jive frame are selected using the information including the number and quality of error bits of the AWZ frame, and the time alignment and the motion vector homogeneity measure are obtained from the block candidate groups. And information on the blocks having the selected high distortion.

The decoding method of the pixel area winner-zeb video codec according to yet another embodiment of the present invention is a decoding method of a pixel area winner-zeb video codec, wherein a previous winner-gib frame transmitted from the encoder is restored, A candidate selection step of selecting a candidate block having a large error rate and a large distortion; a step of calculating a cost for each block in a process of generating supplementary information of a current Winner-Zib frame in a decoder using a temporal correlation for a selected candidate block; A block selection step of selecting a block having a large bit rate and a distortion based on the calculation step and the calculated cost, and an encoding request step of selectively requesting a channel encoding method to the encoder side for the selected block.

Here, the block selection step may be performed in a temporal (temporal) or temporal (temporal) manner by surrounding block information on a block position having a high bit error rate (BER) rate in an AWZ (All Wyner Ziv) frame or a PWZ And blocks are selected based on the correlation.

In the cost calculation step, the cost is calculated by the motion vector homogeneity of the blocks constituting the auxiliary information, and the following formula (6) is used.

In addition, the block selecting step may further include a first sorting step of sorting the time matching cost values of the blocks calculated by Equation (5) in descending order of size, And a second sorting step of sorting the cost values by the method based on the motion vector homogeneity in descending order of magnitude.

In addition, the block selection step may be performed by arranging the order of Ia = 0.4 * It + 0.6 * Im as a sort sequence value, which is Ia = 0.4 * It + 0.6 * Im, A step of rearranging the Ia in order of decreasing size, and a step of selecting L blocks to be transmitted to the encoder in this order.

A computer-readable recording medium according to another embodiment of the present invention is characterized in that a program for executing the encoding and decoding method is recorded.

In the pixel region winner-jive video codec according to another embodiment of the present invention, the previous winner-gib frame transmitted from the encoder side is restored in the pixel region winner-jive video codec, In the process of selecting the block and generating the auxiliary information of the current Winner-Jive frame in the decoder using the temporal correlation with respect to the selected candidate block, the cost is calculated for each block, and the bit rate and the distortion A decoder for selecting a large number of blocks and selectively transmitting a channel coding scheme to the encoder for the selected block and a video frame to be transmitted in accordance with the received channel coding scheme information are divided into an intra picture coding scheme I, (AWZ) considering channel coding, or considering only some blocks as Wiener-jive frames To any one of the channel coding scheme for channel encoding (PWZ) it is characterized in that it comprises a encoder for transmission.

At this time, the decoder decodes temporal correlation through surrounding block information for a block position having a high bit error rate (BER) rate in an AWZ (All Wyner Ziv) frame or a PWZ (Partial Wyner Ziv) And the block is selected based on the block number,

And the time matching cost values of the blocks calculated by Equation (5) are arranged in order of increasing size.

Also, the decoder is characterized in that the cost values by the motion vector homogeneity of the blocks calculated by Equation (6) are sorted in order of magnitude,

It is assumed that the order of the blocks calculated and arranged according to the equations (5) and (6) is It and Im, respectively, and Ia = 0.4 * It + 0.6 * Im. , And selects L blocks to be transmitted to the encoder in this order.

An optional low-power video codec capable of interacting between the encoder and decoder of the present invention having the above-described structure, and a coding and decoding method using the low-power video codec, can selectively improve coding efficiency Effect can be obtained.

In addition, the present invention has the effect of improving the performance of a low-complexity video encoder that can be used in a wireless camera in a near-field environment and a multi-sensor network in a near-field environment.

In addition, the present invention has the effect of enhancing the reliability of images produced for a specific purpose.

Further, the present invention has the effect of improving the bit rate-distortion performance.

FIG. 1 shows two exemplary diagrams of an encoded video sequence according to an embodiment of the present invention.
FIG. 2 shows a structure of an optional low-power video codec device capable of interacting between an encoder and a decoder according to an embodiment of the present invention shown in FIG.
3 is an explanatory diagram of a method of generating auxiliary information in the video codec apparatus of FIG.
FIG. 4 shows a process of selecting block candidates for the current Wiener-Jive frame using the recovered characteristics of the immediately recovered Winner-Jive frame.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, an optional low-power video codec capable of interacting between an encoder and a decoder according to an exemplary embodiment of the present invention will be described in detail with reference to the accompanying drawings. The following drawings are provided by way of example so that those skilled in the art can fully understand the spirit of the present invention. Therefore, the present invention is not limited to the following drawings, but may be embodied in other forms. In addition, like reference numerals designate like elements throughout the specification.

In this case, unless otherwise defined, technical terms and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. In the following description and the accompanying drawings, A description of known functions and configurations that may unnecessarily obscure the description of the present invention will be omitted.

Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings.

1 shows an example of the structure of a coded transmission frame according to an embodiment of the present invention.

First, I represents a frame encoded by the intra picture coding method in the H.264 video coding method. AWZ (All Wyner Ziv) is a method for encoding all constituent blocks of a corresponding transmission frame by using the Winner-Jive frame video coding scheme. The PWZ (Partial Wyner Ziv) indicates that the WiBro-jib frame video coding scheme is used to perform the Wiener-Zib coding for the blocks in the position requested by the decoder, rather than coding all the constituent blocks of the transmission frame. In this case, the period of the I frame is represented by N, and in the embodiment of FIG. 1, N = 2 is shown, which can be changed depending on the application. The period of the AWZ frame is denoted by M, and M = 4 and M = 6 in the embodiment of FIG. 1, respectively.

FIG. 2 shows an implementation of the pixel region Winner-Zib video codec according to the embodiment of FIG.

The pixel region Winner-Zib video codec according to the present invention comprises a PDWZ (Pixel Domain WZ) encoder 10 and a PDWZ (Pixel Domain WZ) decoder 11.

The PDWZ decoder 11 selects a candidate block having a large bit-error rate and distortion by restoring the previous Winner-Zib frame transmitted from the encoder, and uses the temporal correlation for the selected candidate block to decode the current Wiener- In the process of generating the auxiliary information of the jib frame, the cost is calculated for each block, a block having a large bit rate and distortion is selected based on the calculated cost, and a channel encoding method is selectively transmitted to the encoder side for the selected block .

The PDWZ encoder 10 encodes a video frame to be transmitted according to the received channel coding scheme information into an intra picture coding scheme I and an AWZ scheme in which an entire block is regarded as a Wiener- Channel coding in consideration of a jib frame and a channel coding method (PWZ).

The operation of the pixel area winner-jive video codec is described as follows.

As shown in FIG. 2,

First, the PDWZ (Pixel Domain WZ) encoder 10 divides the video frames to be encoded into a key frame and a WZ frame according to the value of N, and processes the key frame by using a conventional intra- (H.264 Intra Encoder) 101, and restores the intra-picture coded image frame and stores it in the first frame memory 102.

At this time, the reconstructed image frame stored in the first frame memory 102 is stored for use as a reference image for the inter-picture difference signal.

The Pixel Domain WZ decoder 11 decodes and restores a bitstream received through an intra-picture decoder (H.264 Intra Decoder) 201. At this time, the restored image information is stored in the second frame memory 104 for decoding and restoring the encoded WZ frame.

The PDWZ decoder 11 generates auxiliary information in the motion compensation interpolating unit 105 using information of the reconstructed key frame stored in the second frame memory 104. [ The operation of the unreliable block predictor 121 is determined depending on whether the next Wiener-jive frame is AWZ or PWZ according to the application of the codec.

The PDWZ encoder 10 can obtain the difference signal by referring to the key frame immediately before the reconstructed WZ frame stored in the first frame memory 102. The first difference signal generator 106 outputs the difference signal r I ask. At this time, if the current frame to be encoded is AWZ, the difference signal r is obtained with respect to the pixels belonging to all the blocks. However, if the frame is PWZ, the difference signal r is obtained with respect to the pixels in the block transmitted from the decoder 11 side. At this time, if the encoding method is a coded block, it is sequentially read and processed by the raster scan method.

At this time, the difference signal r has a Laplacian density distribution characteristic, and performs a bit truncation operation on the difference signal through a first bit (Bit Truncation) In the bit-truncation operation, if the WZ frame and the restored key frame are represented by n bits, these difference signals are represented by (n + 1) bits.

That is, the difference signal r is a signal equal to or greater than -2 ⁿ +1 and equal to or less than 2 ⁿ -1.

For example, when n bits are 8 bits, 2 ⁸ can be expressed from 0 to 255, and the number represented is 256. However, since both negative and positive numbers can be represented, the difference signal r can be expressed from -255 to 255, and the difference signal r is (n + 1) bits.

Since transmission of the entire (n + 1) bits results in waste of a channel, a certain portion of the LSB (Least Significant Bit) is removed and transmitted. In order to increase the efficiency before performing the truncation operation, The difference signal (r) expressed from 255 to 255 is added by 2 ^{< n >} -1 so that all the numbers are positive.

That is, from 255 to 255, 255 is added to the value from 0 to 510.

The remaining difference P is obtained by adding the parameter value (T) according to the transmission bit plane to the difference signal. P can be expressed by the following equation (1).

Where r is the difference between the pixel value of the WZ frame and the restored pixel value of the immediately preceding key frame, n is the number of bits used to express the brightness of the pixel, and T is the parameter value according to the transmission bit plane.

에 대해

으로 나누어 나머지를 취하는 연산에 해당한다.Equation (1)

About

And the rest of the operation.

The remainder obtained in this way can distribute the density distribution variously according to the value of T.

Here, T can be obtained by the following equation (2).

Where k is an integer greater than or equal to 0, n is the number of bits used to represent the pixel, and m is the number of bits discarded for quantization.

Here, the parameter value (T) according to the transmission bit plane is added to the binary code so that the MSB (Most Significant Bit) of the gray code is concentrated in the dense region in consideration of the statistical characteristic.

As described above, in the process of removing a lower bit plane (LSB) of a certain portion, the first quantization unit 108 removes m LSBs and leaves (n + 1-m) MSBs, .

At this time, the gray codes are converted into gray codes through the first gray code unit 109 for the remaining MSBs.

LDPCA (Low-Density Parity-Check Accumulate) encoder 110 sequentially performs channel coding by reading from the MSB of the converted gray code, and stores the generated parity bits in the buffer 111. [

Here, channel coding is one of the most important aspects of information transmission in terms of how much information is transmitted without loss or exactly. In the process of transmitting information, loss or distortion of information is inevitably caused depending on the medium. In order to enable proper transmission of information, it is necessary to transform the information in some way and extract it again by modifying it in a different way.

At this time, channel coding is a combination of a method of transmitting information and a method of receiving and analyzing information.

The parity bit stored in the buffer 111 is gradually transmitted according to the request of the PDWZ decoder 11. [

First, the PDWZ decoder 11 generates auxiliary information by motion compensation interpolation using the key frame reconstructed by the motion compensation interpolator 105 (105). Then, the second-order signal generator 112 generates an inter-frame difference signal frame with respect to the restored key frame immediately before stored in the second frame memory 104 through the generated auxiliary information.

That is, the difference signal r is obtained by the second-order signal generator 112 by comparing the restored key frame stored in the second frame memory 104 with the auxiliary information.

Subsequently, the second bit computing section 113 adds 2 ⁿ -1 to the difference signal r for the inter-picture difference signal r, performs theamplification, and adds the parameter T (113).

In addition, the signal subjected to the bit operation is subjected to quantization through the second quantization unit 114, and gray codes (n + 1-m) of MSB data remaining after the bit operation are quantized using the second Gray code unit 115 .

Then, the LDPCA decoder 116 reads the uppermost (n + 1-m) MSBs with respect to the converted gray code, positions the payload on the received parity bit, and corrects the bit error due to the virtual channel noise.

Here, the payload means a portion including actual information obtained by subtracting a header for various operations and control from a packet or the like.

At this time, if the bit error is not corrected by the received parity bit, the LDPCA decoder 116 requests the buffer 111 of the PDWZ encoder 10 to request additional transmission of a parity bit (Request bits) .

In other words, since the auxiliary information is considered as a form in which noise is added to the information sent from the PDWZ encoder 10, the PDWZ decoder 11 restores the auxiliary information carrying the noise through the received parity bit, and continuously outputs the feedback channel The parity bit is additionally requested to correct the noise.

In this way, the LDPCA decoder 116 completes the correctly corrected reconstruction bit plane through a series of repetitive parity bit requests and bit error correction processes. The restored bit planes are gray codes composed of (n + 1-m) bit planes.

Then, the inverse process for the backward movement and the backward shift is performed on the converted gray code. At this time, the inverse process is performed by the following equation (3).

The inverse mathematical expression for Equation (1) is expressed by Equation (3).

는 부호화된 전체 비트수, T는 전송 비트플레인에 따른 파라미터 값, n은 비트수, 및

은 복원영상에 따른 차이신호를 나타낸다. At this time,

T is the parameter value according to the transmission bit plane, n is the number of bits, and

Represents the difference signal according to the reconstructed image.

는 역비트연산부(119)에 의해 생성된 값이다.
Also,

Is a value generated by the inverse bit computing section (119).

More specifically, the converted gray code is reversely converted back to the binary code through the inverse gray code unit 117. [

At this time, the inverse-transformed binary code is composed of (n + 1-m) bit planes.

Finally, the remaining m LSBs are restored by using the relationship between the restored signal and the auxiliary information through the inverse quantization unit 118. [

If the (n + 1-m) bits are the same as the (n + 1-m) upper bits of the auxiliary information, the LSBs of the auxiliary information m are directly located in the LSB of the recovered signal.

Here, when the LSBs are not equal, the value can be determined such that (n + 1-m) MSBs are located at the boundary value closest to the value of the recovered auxiliary information.

Since the signal generated by Equation (3) is still an inter-picture difference signal, the second-order signal generator 120 adds the restored key frame information stored immediately before in the second frame memory 104 to the Wiener- The restored image is obtained for the frame.

The Winner-Jive frame described above shows a method of coding the same for all blocks. However, when the frame is coded by the PWZ frame, the frame is coded and transmitted to a block that selects the frame and notifies the position of the frame. In this case, a block is selected in the process of generating auxiliary information. FIG. 3 illustrates the function of the motion compensation interpolator 105 for bi-directional motion estimation and compensation for auxiliary information. First, the restored key frame 201 immediately before the time sequence is represented by I _2k-1 , and the restored key frame 202 immediately after the _{rest is} shown by I _{2k + 1} . In the frame 203 to be restored, the current block 204 is obtained by a bi-directional search.

In this calculation formula, SADf (v) is a reconstructed key frame 201 immediately before the current block position or a reconstructed key frame 201 immediately after the block 205 of the winner-zib frame AWZ or PWZ SADb (v) denotes a function expression of the restored key frame 202 immediately after the current block position, which is the same as the current block position, Means a function expression of the block 208 found by searching for the restored key frame 201 or winner-jib frame (AWZ or PWZ) immediately before in the block 206 of the winner-gib frame (AWZ or PWZ). Here, SAD stands for Sum of Absolute Difference. In these two relations, M represents pixels included in the blocks 205 and 206 which are the starting points, v is the motion vector candidates, and v, which is the minimum value in the equation (4), is determined by the motion vector v *.

Thus, when all blocks are filled with motion compensation interpolation, the generation of auxiliary information is completed. In this case, the compensation interpolation performance is indirectly evaluated for each block by two methods.

First, with respect to the block obtained by Equation (4), the temporal matching cost is measured with respect to a region in which the size of the motion vector is reduced by half in the bidirectional frame subjected to motion compensation interpolation. Equation 5 is used for the time matching cost.

In this way, the time alignment cost values are obtained for all the blocks, and the order is sorted in the order of larger size. When sorted, the sort order is listed for each block number.

The second measure of cost is cost by motion vector homogeneity, but Equation (6) is used.

Here, S denotes surrounding blocks surrounding the current block, and w denotes a motion vector of the blocks. That is, Equation (6) measures the distance between the motion vector of the current block and the motion vector between the neighboring blocks, and measures the smallest value as a cost value. The cost thus obtained exists as many as the number of blocks, and this is also arranged in the order of size as the time matching cost.

FIG. 4 shows a process of selecting block candidates for the current Wiener-Jive frame using the recovered characteristics of the immediately recovered Winner-Jive frame.

The MSB bit error rate can be calculated in the previously recovered AWZ frame or PWZ frame and the candidate block can be designated for the current frame as shown in FIG. (5) and (6), and the results are sorted in order of magnitude, this is sorted from 1 to the number of blocks (L). In this case, an index value of 1, which is the largest, is assigned, a value corresponding to the sorted order is assigned to each block, and an index L is assigned to the block having the smallest value. When the index values assigned as described above are represented by I _t and I _m through Equation (7), the combined index value I _a can be derived.

In this way, the combined index value is obtained for each index value, and each index value is obtained and rearranged for the result (Ia). At this time, if an equivalent value is determined, let the value of I _t be large in the sort order.

For each of the blocks thus obtained, the upper P block positions are transmitted to the encoder to request reception. At this time, a coding method such as an Exp-Golomb code can be used as a coding method for P block positions.

Hereinafter, a coding method and a decoding method of the pixel-area Winner-Jive video codec according to the present invention will be described.

The coding method of the pixel region Winner-Zib video codec according to the present invention receives channel coding scheme information determined based on the cost information calculated from the auxiliary information generated by the decoder. In addition, a method (AWZ) in which a video frame to be transmitted is considered as an intra picture coding scheme (I) and an entire block is regarded as a Wiener-jive frame according to the received channel coding scheme information, or a method Channel coding method (PWZ), and transmits the resultant. In the encoding method of the pixel area Wiener-Jive video codec, the encoder encodes only a part of the blocks as a Wiener-Jive frame and then transmits the channel coding scheme information. The channel coding scheme information includes a Wiener- ), The number of error bits and the quality of the error bits of the current winner-jib frame, and selects the block candidate groups for the current winner-jib frame, obtains the time alignment and the motion vector homogeneity measure from the block candidate groups, Lt; / RTI >

The decoding method of the pixel area winner-zeb video codec according to the present invention decodes the previous Winner-Zib frame transmitted from the encoder and selects a candidate block having a large bit-error rate and a large distortion, In the process of generating auxiliary information of the current Winner-Jive frame in the decoder, a cost is calculated for each block, a block having a large bit rate and distortion is selected based on the calculated cost, And selectively requests the channel coding method. The decoding method of the pixel area winner-zip video codec is a method of decoding a video signal of a surrounding area for a block position having a high BER (Bit Error Rate) rate in an AWZ (All Wyner Ziv) frame or a PWZ A block is selected based on temporal correlation through block information. The costs are calculated as the cost of the time alignment of the blocks constituting the auxiliary information, and the cost is calculated by the motion vector homogeneity of the blocks constituting the auxiliary information. The temporal matching cost values of the calculated blocks are sorted in descending order of magnitude and the cost values by the motion vector homogeneity of the calculated blocks are sorted in ascending order of magnitude. The order of each ordered block is combined and rearranged in order of size, and the block position information is transmitted to the encoder by selecting L blocks to be transmitted to the encoder in the rearranged order.

In addition, although the method described in the present invention is described through the pixel region Winner-Zib coding method, the process of applying the DCT transform region can be used as it is. Therefore, the present invention has the effect of improving the performance of a video encoder of a more complicated complexity that can be used in a mobile terminal, a wireless camera, a sensor, and the like. The present invention increases the generation of user- .

Meanwhile, the pixel region Winner-Zib encoding and decoding method according to the embodiment of the present invention may be implemented in a form of a program command which can be performed through various electronic information processing means and recorded in a storage medium. The storage medium may include program instructions, data files, data structures, and the like, alone or in combination.

Program instructions to be recorded on the storage medium may be those specially designed and constructed for the present invention or may be available to those skilled in the art of software. Examples of storage media include magnetic media such as hard disks, floppy disks and magnetic tape, optical media such as CD-ROMs and DVDs, magneto-optical media such as floptical disks, magneto-optical media and hardware devices specifically configured to store and execute program instructions such as ROM, RAM, flash memory, and the like. The above-mentioned medium may also be a transmission medium such as a light or metal wire, wave guide, etc., including a carrier wave for transmitting a signal designating a program command, a data structure and the like. Examples of program instructions include machine language code such as those produced by a compiler, as well as devices for processing information electronically using an interpreter or the like, for example, a high-level language code that can be executed by a computer.

As described above, the present invention has been described with reference to specific embodiments such as specific components and exemplary embodiments. However, the present invention is not limited to the above-described embodiments, And various modifications and changes may be made thereto by those skilled in the art to which the present invention pertains.

Accordingly, the spirit of the present invention should not be construed as being limited to the embodiments described, and all of the equivalents or equivalents of the claims, as well as the following claims, fall within the scope of the present invention .

10: Pixel Domain WZ (PDWZ) encoder
11: PDWZ (Pixel Domain WZ) decoder
101: H.264 intra coder 103: H.264 intra decoder
102: first frame memory 104: second frame memory
105: Motion compensation interpolator 106: Primary signal generator
107: first bit computing unit 108: first quantization unit
109: first gray code section
110: LDPCA (Low-Density Parity-Check Accumulate) encoder
111: buffer 112: second primary signal generator
113: second bit computing unit 114: second quantization unit
115: second Gray code part
116: LDPCA (Low-Density Parity-Check Accumulate) decoder
117: reverse gray code unit 118: dequantization unit
119: Inverse bit computing unit 120: Second-
121: Unreliable block predictor

Claims (13)

A method of encoding a pixel area winner-jive video codec,
Receiving channel coding scheme information determined based on the cost information calculated from the auxiliary information generated by the decoder; And
(AWZ) in which a video frame to be transmitted is considered as an intra picture coding scheme (I) and an entire block is regarded as a Wiener-jive frame according to the received channel coding scheme information, or a method (PWZ), and transmitting the channel-encoded data;
The method of claim 1, wherein the encoding of the pixel region of the pixel region is performed by the encoding means.
The method according to claim 1,
The encoding method of the pixel area winner-zeb video codec
The encoder encodes only some of the blocks as a Wiener-Jive frame and performs channel coding and transmits the information including the number and quality of erroneous bits of the winner-jib frame AWZ received immediately before the decoder to the channel coding scheme information The motion vector homogeneity measure is obtained from among the block candidate groups and the information about the blocks having the highest distortion selected by the combination of these measures is included And encoding the pixel region winner-zeb video codec.
A decoding method of a pixel area winner-jive video codec,
A candidate selecting step of selecting a candidate block having the largest bit-error rate and distortion by restoring the previous Winner-Jib frame transmitted from the encoder;
A cost calculation step of calculating a cost for each block in a process of generating supplementary information of a current Winner-Jive frame in a decoder using a temporal correlation for a selected candidate block;
A block selecting step of selecting a block having the largest bit rate and distortion based on the calculated cost; And
An encoding request step of selectively requesting a channel encoding method on a selected block for a coder;
The method of claim 1, further comprising:
The method of claim 3,
The block selection step
Block based on temporal correlation through surrounding block information for a block position having the highest BER (Bit Error Rate) rate in an AWZ (All Wyner Ziv) frame or a PWZ (Partial Wyner Ziv) frame received and restored just before And selecting the pixel region winner-zeb video codec.
The method of claim 3,
The cost calculation step
Wherein the cost is calculated by a method of motion vector homogeneity of blocks constituting auxiliary information, and the following equation is used.

6. The method of claim 5,
The block selection step
A first sorting step of sorting the time matching cost values of the blocks calculated by the following equations in descending order of magnitude;
The method of claim 1, further comprising:

The method of claim 6, wherein
The block selection step
A second sorting step of sorting the cost values according to the motion vector homogeneity of the blocks calculated by the following equations in descending order of magnitude;
The method of claim 1, further comprising:

8. The method of claim 7,
The block selection step
Replacing Ia = 0.4 * It + 0.6 * Im with an alignment order value merged into Ia = 0.4 * It + 0.6 * Im, wherein the order of each block aligned in the first and second alignment steps is It and Im respectively;
Rearranging Ia in order of decreasing size; And
Selecting L blocks to be transmitted to the encoder in this order;
The method of claim 1, further comprising:
In a pixel region winner-jive video codec,
A candidate block having the largest bit error rate and distortion is selected by restoring the previous Winner-Zible frame transmitted from the encoder side, and auxiliary information of the current Winner-Jive frame is decoded in the decoder using the temporal correlation for the selected candidate block A decoder for selecting a block having the largest bit rate and distortion based on the calculated cost and selectively transmitting a channel coding scheme to a coder side for a selected block; And
(AWZ) in which a video frame to be transmitted is considered as an intra picture coding scheme (I) and an entire block is regarded as a Wiener-jive frame according to the received channel coding scheme information, or a method (PWZ), and transmits the channel-encoded signal;
Wherein the pixel region winner-zip video codec comprises:
10. The method of claim 9,
The decoder
Block based on temporal correlation through surrounding block information for a block position having the highest BER (Bit Error Rate) rate in an AWZ (All Wyner Ziv) frame or a PWZ (Partial Wyner Ziv) frame received and restored just before Pixel video codec.
10. The method of claim 9,
The decoder
Wherein the time alignment cost values of blocks calculated by the following equations are sorted in descending order of magnitude.

12. The method of claim 11,
The decoder
Wherein a cost value by a method of motion vector homogeneity of blocks calculated by the following equation is sorted in descending order of magnitude.

13. The method of claim 12,
The decoder
It is assumed that Ia = 0.4 * It + 0.6 * Im, where Ia and Im are the order of each block calculated and arranged according to the following equations: Ia = And selects L blocks to be transmitted to the encoder in this order.

Images