CN1206299A - Method and apparatus for contour motion estimating binary image by using weighted block match algorithm - Google Patents
Method and apparatus for contour motion estimating binary image by using weighted block match algorithm Download PDFInfo
- Publication number
- CN1206299A CN1206299A CN 97116879 CN97116879A CN1206299A CN 1206299 A CN1206299 A CN 1206299A CN 97116879 CN97116879 CN 97116879 CN 97116879 A CN97116879 A CN 97116879A CN 1206299 A CN1206299 A CN 1206299A
- Authority
- CN
- China
- Prior art keywords
- pixel
- search block
- border
- weight
- candidate blocks
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Landscapes
- Compression Or Coding Systems Of Tv Signals (AREA)
Abstract
A motion estimating method, for use in a motion estimation between a current and a previous frames of a binary video image, wherein the current frame is divided into a plurality of equal sized search blocks, forms a plurality of candidate blocks, each of the candidate blocks having a same size as the search block, to thereby provide a displacement of each candidate block from the search block as a displacement vector of each candidate block. Thereafter, the method overlaps the search block with each candidate block to thereby determine an unmatched region, a boundary of a matched region and a boundary of the search block and assigns a weight to each of the pixels within the unmatched region on a predetermined basis, wherein the weight may vary from pixel to pixel. Then, the method calculates a weighted error of each candidate block by using the weights and selects a displacement vector which has a minimum error as an optimum motion vector.
Description
The present invention relates to a kind of being used for by using a weighting block matching algorithm to find out the method and apparatus of an optimum movement vector that produces minimum weighted; And, more specifically, relate to a kind of being used for by distributing a weight that changes with the pixel of a frame of video to determine the method and apparatus of weighted error.
The transmission of digital video signal can be transmitted than the much higher video image of quality with the video image of analog signal transmission.When a picture intelligence that comprises a sequence chart picture frame is represented with a digital form, just produced lot of data, especially under the situation of high-definition television system.Yet, because the usable frequency limited bandwidth of conventional transmission channel in order to send these a large amount of numerical datas by this limited channel width, just needs compression or reduces transmitted data amount.In various video compression technologies, so-called mixed coding techniques is considered to the most effective, and this mixed coding technology combines time and spatial compression techniques together and statistical coding technology.
The mixed coding techniques of great majority adopts a motion compensation DPCM (differential pulse-code modulation), two-dimensional dct (discrete cosine transform), DCT coefficient quantization, and VLC (variable length code).Motion compensation DPCM is the definite motion of a target between present frame and its reference frame (a for example previous frame), according to the motion stream prediction present frame of this target, and produces the process that the differential signal of the difference between preceding frame and its predicted value is deserved in a representative.
The two-dimensional dct of minimizing or removal such as the spatial redundancy between the pictorial data of motion compensation DPCM data with a digital image data piece, for example one of 8 * 8 pixels, converts one group of DCT coefficient data to.This technology is described in Chen and Pratt's on " Scene Adaptive Coder " (scene adaptive coding device), publishes in IEEE Transctions on Communications, COM-32, NO.3,225-231 page or leaf (in March, 1984).By handle this PCT coefficient data with a quantizer, zigzag scanner and VLC, data volume to be sent out can be effectively incompressible.
Especially, in motion compensation DPCM, for example by using block matching algorithm, according to estimation to the motion between present frame and the previous frame, current frame data is doped from previous frame data (sees that for example people such as J.R.J ain " Displacement Measurement and Its Application inInterframe Image Coding " (displacement measurement and the application the interframe picture coding thereof) publishes in IEEE Transactions of Communications, COM-29, NO.12,1799-1808 page or leaf (in December, 1981)).This estimated motion can be described according to the two-dimensional motion vector of the displacement of expression pixel between present frame and previous frame.
According to block matching algorithm, a present frame is divided into a plurality of search block.The size of one search block is the scope between 8 * 8 and 32 * 32 pixels typically.Be in a present frame, to determine the motion vector of a search block, carry out a similitude between each in the bigger usually field of search in the search block of this present frame and a previous frame in the candidate blocks of included a plurality of equal sizes and calculate.One error function of use such as mean absolute error or mean square deviation is carried out this search block and the similitude between each candidate blocks in this field of search of this present frame and is calculated.And by definition, a motion vector has been represented this search block and has been produced displacement between the candidate blocks of a smallest error function.
As everyone knows, each pixel in the binary image has the sign in its affiliated district of identification: for example, the pixel in a background is identified as " 0 "; And the pixel in a target is identified as non-0 value.Therefore, if a contour motion that will use conventional block matching algorithm as the mean absolute error of error function to be used for binary image is estimated, then the number that does not mate pixel is just counted in the effect of this error function, wherein each not mate pixel be a pixel in the candidate blocks, the corresponding pixel in a search block of this candidate blocks has the ident value different with it.
With reference to figure 1, show a search block SB, one first candidate blocks CB1 and one second candidate blocks CB2 have the not coupling pixel that is designated as " 0 ".As illustration, when the search block SB and the first candidate blocks CB1 overlaid, not mating number of picture elements is 6; And when the search block SB and the second candidate blocks CB2 overlaid, not mating number of picture elements is 7.So the first candidate blocks CB1 is selected as the piece that produces smallest error function, and the first candidate blocks CB1 is selected as optimum movement vector apart from the displacement of search block SB, though the second candidate blocks CB2 than the first candidate blocks CB1 more similar to search block SB in shape.
If select motion vector in this way, then the real effectively profile information in determining a target shape may in the shape of determining a target, can not play a part suitable
Therefore, a main purpose of the present invention provides a kind of being used for by using a weighting block matching algorithm to find out the method and apparatus of the optimum movement vector of a generation minimum weighted.
According to the present invention, a kind of method is provided, be used to carry out the present frame of binary video image and the estimation between the previous frame, wherein present frame is divided into a plurality of equal-sized search block, this method comprises the following steps: a) formerly to form a plurality of fields of search in the frame, and wherein each field of search is corresponding to each search block; B) form a plurality of candidate blocks in each field of search, each candidate blocks has the size identical with search block, thereby each candidate blocks is provided as the displacement vector of each candidate blocks apart from the displacement of this search block; C), thereby determine a border that coupling is not distinguished, a coupling is distinguished and the border of this search block with overlapping this search block of each candidate blocks; D) distribute a weight to give not each interior pixel of coupling district; E) in step (d), distribute to the weighted error that the weight of mating each pixel in the district is calculated each candidate blocks by utilizing; F) these weighted errors are done mutually relatively, thereby selected minimum weighted, and provide one to select signal, wherein which displacement vector this selection signal indicates corresponding to minimum weighted; And g) selects signal in response to this, select displacement vector, thereby it is provided as optimum movement vector corresponding to minimum weighted.
Above and other purpose of the present invention and characteristics will become obviously by description of preferred embodiments with reference to the accompanying drawings, in the accompanying drawing:
Fig. 1 shows a kind of method that is used for determining by a conventional block matching algorithm error function between a search block and the candidate blocks;
Fig. 2 represents the block diagram that includes a contour encoding device of motion estimation unit according to of the present invention;
Fig. 3 A and 3B have illustrated a detailed view of motion estimation unit shown in Figure 2;
Fig. 4 A has illustrated a kind of edge pixel that is used for according to the coupling district according to the first embodiment of the present invention, the method that assigns weight; And
Fig. 4 B has illustrated that according to a second embodiment of the present invention a kind of is used for the method that the edge pixel according to search block assigns weight.
With reference to figure 2, the block diagram that includes the contour encoding device 1 of a motion estimation unit according to of the present invention has been described.
The binary picture data of present frame, i.e. current frame data is provided for a search block with the form of segmentation mask (segmentatim mask) and forms unit 100, and wherein each pixel in this segmentation mask has the sign in its zone that belongs to of identification.For example, the pixel in the background is identified as " 0 " and each pixel in the target is identified as non-0 value.Search block forms unit 100 this present frame is divided into a plurality of equal-sized search block, thereby provides the search block data by circuit L10 to subtrator 200 and motion estimation unit 400.
Deduct the optimal candidate piece of the motion compensation on the circuit L30 the search block of subtrator 200 from circuit L10, thereby provide a subtraction result, be i.e. an error signal to encoder 600.
700 pairs of these encoded error signals of decoder are decoded and should decoded error signal be sent to adder unit 800.Adder unit 800 adds to the best candidate piece of the motion compensation on the circuit L30 with this decoded error signal, thereby produces a reconstruction signal of this present frame.This reconstruction signal of this present frame is stored in the frame memory 300.
With reference to figure 3A and Fig. 3 B, show the more detailed block diagram of motion estimation unit shown in Figure 2 400.
Frame data is offered the field of search and is formed unit 410 from frame memory 300 via line L20.The field of search form unit 410 usefulness a certain size, shape and search pattern define and the corresponding field of search of this search block, carries out the estimation of search block thus.After the field of search had been determined in formation unit 410, the field of search, these field of search data then were provided for candidate blocks and form unit 420-1 to 420-N, and N is the positive integer of the sum of the formed candidate blocks of a representative.
Have a plurality of candidate blocks and form the unit, yet for simplicity, only described 3 unit.Form unit 420-1 to 420-N in each candidate blocks, in this field of search, produce a candidate blocks with the identical size of this search block; And the pixel data of each candidate blocks is exported to each piece matching unit 430-1 to 430-N thus.These candidate blocks also are used as displacement vector DV1 to DV apart from the relative displacement of the position of the search block of present frame
NForm unit 420-1 to 420-N from candidate blocks and export to a multiplexer (MUX) 470.
As mentioned above, the search block data from search block formation unit 100 on the circuit L10 also are provided for piece matching unit 430-1 to 430-N.At each piece matching unit 430-1 to 430-N, the pixel value of the pixel value of search block and corresponding candidate piece is compared mutually.Especially, wherein corresponding candidate blocks and this search block overlaid of each piece matching unit 430-1 to 430-N; And make upward sign for the edge pixel in the edge pixel of this search block on the search block data, the pixel that does not mate the district and coupling district.
Because search block data and candidate blocks data all are the segmentation mask forms, if therefore this pixel belongs to background, then this pixel value is " 0 "; This pixel value is if this pixel belongs to a target, for example is 1.The edge pixel of this search block is a pixel in this search block, and its pixel value is " 1 ", and the pixel value of the pixel that is adjacent is " 0 "; The pixel of matching area is not a pixel in this search block, and the corresponding pixel of this pixel in the corresponding candidate piece has the pixel value different with it; And the edge pixel in this coupling district is the pixel in this search block, and this pixel is adjacent with a pixel that does not mate the district or adjacent for a pixel of " 0 " with pixel value when its pixel value is " 1 ".
This search block data via line L50-1 to L50-N that pixel place at the pixel place on the border of this search block, in coupling district not and the quilt at the edge pixel place that coupling is distinguished make a check mark is provided for weight allocation unit 440-1 to 440-N respectively.
According to the first embodiment of the present invention, each weight allocation unit 440-1 to 440-N does not mate pixel in the corresponding candidate piece each to distribute a weight, wherein these not mate pixel be this not pixel in the matching area, and degree of closeness of the edge pixel that this weight is distinguished according to itself and coupling does not change with mating the different of pixel.The weight that the nearer quilt that does not mate pixel of edge pixel from this coupling district adds is less than the edge pixel weight of not mating pixel far away from this coupling district.In other words, be positioned near one the edge pixel in coupling district and do not mate pixel than far away one not the mating pixel and be assigned with a less weight of the edge pixel from this coupling district.
With reference to figure 4A, the exemplary plot according to the first embodiment of the present invention is shown.SBi (i is 1 or 2) indication the edge pixel place sign in a coupling district and do not mate the painted search block data in place, district one, as with search block SB shown in Figure 1 and i the result that candidate blocks CBi makes comparisons shown in Figure 1.The pixel of black is that the border in this coupling district resembles rope; Adjacent each of edge pixel in coupling district is not mated pixel and has been assigned with W1 therewith; Each pixel that does not mate that leaves a pixel place of edge pixel in this coupling district is assigned W2; Each pixel that leaves 2 pixel places of edge pixel in this coupling district has been assigned with W3, wherein W1 be one on the occasion of, W2 is greater than W1, W3 is greater than W2.
According to a second embodiment of the present invention, on the other hand, each weight allocation unit 440-1 to 440-N does not mate pixel in the corresponding candidate piece each to distribute a weight, and wherein this weight does not change with mating the different of pixel with the degree of closeness of the edge pixel of this search block according to mating pixel.From the edge pixel of this search block nearer do not mate pixel be assigned with than from the edge pixel of this search block far away do not mate the less weight of pixel.In other words, be positioned at and do not mate pixel with one of the edge pixel same position place of this search block and be assigned with minimal weight; And the edge pixel of leaving this search block pixel that do not mate far away is assigned with bigger weight.
With reference to figure 4B, show an exemplary diagram according to a second embodiment of the present invention.SBi (i is 1 or 2) also indicate the edge pixel place in coupling district sign and do not mate the painted search block data in place, district one, as with search block SB shown in Figure 1 and i the result that candidate blocks CBi makes comparisons shown in Figure 1.The pixel of black is the edge pixel of this search block; Be positioned at and do not mate pixel with each of the edge pixel same position in this coupling district and be assigned with W1; Each adjacent with the pixel in this coupling district is not mated pixel and has been assigned with W2; And each of a pixel place of pixel of leaving this coupling district is not mated pixel and has been assigned with W3; And each of two pixel places of pixel of leaving this coupling district do not mate pixel and has been assigned with W4, wherein W1 be one on the occasion of, W2 is greater than W1, W3 is greater than W2, W4 is greater than W3.
It does not mate the search block data that pixel is weighted and is provided for weighted error computing unit 450-1 to 450-N.Each weighted error computing unit 450-1 to 450-N calculates the weighted error of corresponding candidate piece.Especially, at each weighted error computing unit 450-1 to 450-N, be assigned to all weighted values that do not mate pixel in the corresponding candidate piece and added up to produce weighted error, the weighted error of the corresponding candidate piece that calculates is provided for comparator 460.
Back with reference to figure 4, will describe according to the first embodiment of the present invention how to calculate weighted error now.Suppose that W1 is 1, W2 is 2, and W3 is 3, and the not coupling number of picture elements of SB1 is 6, and the weighted error WE1 that calculates (SB1) is 12, and is as follows:
WE1(SB1)=W3+W2+W1+W1+W2+W3
=2 (W1+W2+W3)=12 and, the not coupling number of picture elements of SB2 is 7, the weighted error WE1 that then calculates (SB2) is 7, and is as follows:
WE1(SB2)=W1+W1+W1+W1+W1+W1+W1
=7W1=7
With reference to figure 4B, it can illustrate how to calculate weighted error according to second embodiment of the invention.Suppose that W1 is 1, W2 is 2, and W3 is 3, and W4 is 4, and the not coupling number of picture elements of SB1 is 6, and the weighted error WE2 that calculates (SB1) is 12, and is as follows:
WE2(SB1)=W2+W3+W4+W1+W1+W1
=3W1+W2+W3+W4=12 and, the not coupling number of picture elements of SB2 is 7, the weighted error WE2 that then calculates (SB2) is 8, and is as follows:
WE2(SB2)=W1+W1+W1+W1+W1+W2+W1
=6W1+W2=8
By calculating weighted error according to the first and second aspects of the present invention, has minimum weighted corresponding to the second candidate blocks CB2 of SB2.
Simultaneously, comparator 460 compares weighted error mutually, thereby selects a minimum weighted, and provides one to select signal, and this signal makes multiplexer 470 can select displacement vector corresponding to this minimum weighted.
As implied above, by utilizing the weighting block matching algorithm, a contour motion is more effectively estimated according to the present invention.
Though the description of this invention carries out according to specific embodiment,, show under the prerequisite of the spirit and scope that do not deviate from the claims definition and can make variations and modifications the present invention for those skilled in the art.
Claims (10)
1, a kind of method, be used to carry out the present frame of binary image signal and the estimation between the previous frame, wherein present frame is divided into a plurality of equal-sized search block, this method comprises the following steps: a) formerly to form a plurality of fields of search in the frame, and wherein each field of search is corresponding to each search block; B) form a plurality of candidate blocks in each field of search, each candidate blocks has the size identical with search block, thereby each candidate blocks is provided as the displacement vector of each candidate blocks apart from the displacement of this search block; C), thereby determine a border that coupling is not distinguished, a coupling is distinguished and the border of this search block with overlapping this search block of each candidate blocks; D) distribute a weight to give not each interior pixel of coupling district; E) in step (d), distribute to the weighted error that the weight of mating each pixel in the district is calculated each candidate blocks by utilizing; F) these weighted errors are done mutually relatively, thereby selected minimum weighted, and provide one to select signal, wherein which displacement vector this selection signal indicates corresponding to minimum weighted; And g) selects signal in response to this, select to be provided as optimum movement vector corresponding to the displacement vector of minimum weighted and with it.
2, according to the process of claim 1 wherein that described step (c) may further comprise the steps:
(c1) with each candidate blocks and this search block overlaid;
(c2) each pixel value of this search block and the corresponding pixel value of this candidate blocks are made comparisons, wherein the corresponding pixel value of this candidate blocks is to be arranged in and value with the pixel at its this search block one pixel same position place that compares; And
(c3) border of distinguishing, the border that reaches search block are distinguished, mated to definite coupling, wherein the not coupling district in this search block is made up of a plurality of pixels, the corresponding pixel of these pixels in the corresponding candidate piece has the pixel value different with it, the border in the coupling district in this search block by its pixel value be 1 and adjacent with coupling district not or with pixel value be that 0 pixel neighboring pixels is formed, and the border of this search block by its pixel value be 1 and with pixel value be that 0 pixel neighboring pixels is formed.
3, according to the method for claim 2, wherein said step (e) is carried out phase Calais calculating weighted error by all being distributed to this weight of not mating the pixel in the district.
4, according to the method for claim 3, wherein said step (d) is distributed a weight according to this pixel and the degree of closeness on the border in coupling district to not distinguishing each pixel of joining in the district, so that, less weight is given from the nearer pixel in the border in this coupling district, and bigger weight is given from the border in this coupling district pixel far away.
5, according to the method for claim 3, wherein said step (d) is given not according to the degree of closeness on the border of this pixel and search block, and each interior pixel of coupling district distributes a weight, so that, less weight is given the pixel of the boundary that is positioned at this search block, and bigger weight is given the pixel on the border of leaving this search block.
6, a kind of device is used to carry out the present frame of binary image signal and the estimation between the previous frame, and wherein present frame is divided into a plurality of equal-sized search block, and this device comprises:
One field of search forms device, is used for formerly forming a plurality of fields of search in the frame, and wherein each field of search is corresponding to each search block; One candidate blocks forms device, is used for forming a plurality of candidate blocks in each field of search, and each candidate blocks has the size identical with search block, thereby each candidate blocks is provided as the displacement vector of each candidate blocks apart from the displacement of this search block;
A coalignment is used for overlapping this search block of each candidate blocks, thereby determines a border that coupling is not distinguished, a coupling is distinguished and the border of this search block;
One weight allocation device is used to distribute a weight to give not each interior pixel of coupling district;
One weighted error calculation element is used for distributing to the weighted error that the weight of mating each pixel in the district is calculated each candidate blocks by utilizing at the weight allocation device;
One comparison means is used for these weighted errors are done mutually relatively, thereby selects minimum weighted, and provides one to select signal, and wherein which displacement vector this selection signal indicates corresponding to minimum weighted;
One multiplex machine is used in response to this selection signal, selects to be provided as optimum movement vector corresponding to the displacement vector of minimum weighted and with it.
7, according to the device of right 6, wherein said coalignment comprises:
One overlapping device is used for each candidate blocks and this search block overlaid;
One coalignment is used for each pixel value of this search block and the corresponding pixel value of this candidate blocks are made comparisons, and wherein the corresponding pixel value of this candidate blocks is to be arranged in and value with the pixel at its this search block one pixel same position place that compares;
One determines device, be used for definite coupling district, mate the border of distinguishing, the border that reaches search block, wherein the not coupling district in this search block is made up of a plurality of pixels, the corresponding pixel of these pixels in the corresponding candidate piece has the pixel value different with it, the border in the coupling district in this search block by its pixel value be 1 and adjacent with coupling district not or with pixel value be that 0 pixel neighboring pixels is formed, and the border of this search block by its pixel value be 1 and with pixel value be that 0 pixel neighboring pixels is formed.
8, according to the method for claim 7, wherein said weighted error calculation element carries out phase Calais calculating weighted error by all being distributed to this weight of not mating the pixel in the district.
9, method according to Claim 8, wherein said weight allocation device distributes a weight according to this pixel and the degree of closeness on the border in coupling district to not distinguishing each pixel of joining in the district, so that, less weight is given from the nearer pixel in the border in this coupling district, and bigger weight is given from the border in this coupling district pixel far away.
10, method according to Claim 8, wherein said weight allocation device is given not according to the degree of closeness on the border of this pixel and search block, and each interior pixel of coupling district distributes a weight, so that, less weight is given the pixel of the boundary that is positioned at this search block, and bigger weight is given the pixel on the border of leaving this search block.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 97116879 CN1206299A (en) | 1997-07-19 | 1997-09-03 | Method and apparatus for contour motion estimating binary image by using weighted block match algorithm |
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| KR33820/97 | 1997-07-19 | ||
| CN 97116879 CN1206299A (en) | 1997-07-19 | 1997-09-03 | Method and apparatus for contour motion estimating binary image by using weighted block match algorithm |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN1206299A true CN1206299A (en) | 1999-01-27 |
Family
ID=5174172
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN 97116879 Pending CN1206299A (en) | 1997-07-19 | 1997-09-03 | Method and apparatus for contour motion estimating binary image by using weighted block match algorithm |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN1206299A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN100385936C (en) * | 2003-01-23 | 2008-04-30 | 皇家飞利浦电子股份有限公司 | Background motion vector detection |
| CN102587591A (en) * | 2012-03-19 | 2012-07-18 | 王�忠 | Light roofing system |
-
1997
- 1997-09-03 CN CN 97116879 patent/CN1206299A/en active Pending
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN100385936C (en) * | 2003-01-23 | 2008-04-30 | 皇家飞利浦电子股份有限公司 | Background motion vector detection |
| CN102587591A (en) * | 2012-03-19 | 2012-07-18 | 王�忠 | Light roofing system |
| US9556616B2 (en) | 2012-03-19 | 2017-01-31 | Hainan Sequoia Creation Co., Ltd. | Light roofing system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN1146241C (en) | Method and apparatus for encoding video signal of contour of object | |
| CN100473167C (en) | Method for encoding images, and image coder | |
| CN1117482C (en) | Method for encoding video signal using feature point based motion estimation | |
| US5969766A (en) | Method and apparatus for contour motion estimating a binary image by using a weighted block match algorithm | |
| CN1134090A (en) | Method and apparatus for encoding/decoding video signal | |
| KR100488422B1 (en) | Grayscale-shaped information encoding / decoding device and method | |
| US20020090030A1 (en) | Digital image encoding and decoding method and digital image encoding and decoding device using the same | |
| CN1166211C (en) | Method and apparatus for encoding motion vector based on number of vilid reference motion vectors | |
| CN1131872A (en) | Image processing system using pixel-by-pixel motion estimation and frame decimation | |
| GB2320836A (en) | Error concealing in video signal decoding system | |
| CN1208313A (en) | Method and apparatus for encoding motion vector | |
| CN1103164C (en) | An image processing system using pixel-by-pixel motion estimation and frame decimation | |
| CN1136732C (en) | Improved motion compensation apparatus for use in image encoding system | |
| CN1111336C (en) | Current image predictive method and device in image signal coding system | |
| US5883977A (en) | Method and apparatus for encoding a video signal of a contour of an object | |
| CN102017626A (en) | Method of coding, decoding, coder and decoder | |
| CN1136728C (en) | Method and equipment for testing optimum movement vector | |
| CN1132984A (en) | Block matching motion estimation method | |
| KR100212559B1 (en) | The contour coding system and motion prediction method for object | |
| GB2320989A (en) | Padding image blocks using substitute pixels for shape adaptive discrete cosine transform | |
| CN1206299A (en) | Method and apparatus for contour motion estimating binary image by using weighted block match algorithm | |
| CN1078795C (en) | Improved motion compensation method for use in image encoding system | |
| CN1142733A (en) | Image processing system using feature point-based motion estimation | |
| KR100220680B1 (en) | The vertex coding apparatus for object contour encoder | |
| KR100220581B1 (en) | The vertex coding apparatus for object contour encoder |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C01 | Deemed withdrawal of patent application (patent law 1993) | ||
| WD01 | Invention patent application deemed withdrawn after publication |