CN1216498C - Sub-sample interpolation filter method - Google Patents
Sub-sample interpolation filter method Download PDFInfo
- Publication number
- CN1216498C CN1216498C CN 200310101370 CN200310101370A CN1216498C CN 1216498 C CN1216498 C CN 1216498C CN 200310101370 CN200310101370 CN 200310101370 CN 200310101370 A CN200310101370 A CN 200310101370A CN 1216498 C CN1216498 C CN 1216498C
- Authority
- CN
- China
- Prior art keywords
- sample
- whole
- interpolation
- filter
- row
- 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.)
- Expired - Lifetime
Links
- 238000000034 method Methods 0.000 title claims abstract description 57
- 238000001914 filtration Methods 0.000 claims abstract description 26
- 239000000523 sample Substances 0.000 claims description 226
- 238000003780 insertion Methods 0.000 claims description 15
- 230000037431 insertion Effects 0.000 claims description 15
- 238000012935 Averaging Methods 0.000 claims description 13
- 238000012545 processing Methods 0.000 claims description 9
- 239000013074 reference sample Substances 0.000 claims description 3
- 238000013461 design Methods 0.000 abstract description 2
- 101150039623 Clip1 gene Proteins 0.000 description 12
- 229910052739 hydrogen Inorganic materials 0.000 description 3
- 241000023320 Luma <angiosperm> Species 0.000 description 2
- OSWPMRLSEDHDFF-UHFFFAOYSA-N methyl salicylate Chemical compound COC(=O)C1=CC=CC=C1O OSWPMRLSEDHDFF-UHFFFAOYSA-N 0.000 description 2
- 101000969688 Homo sapiens Macrophage-expressed gene 1 protein Proteins 0.000 description 1
- 102100021285 Macrophage-expressed gene 1 protein Human genes 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000002123 temporal effect Effects 0.000 description 1
Abstract
The present invention relates to a separate sample interpolation filter method which belongs to the technical field of encoding and decoding videos and images. The method is used for carrying out 1/2 sample interpolation and 1/4 sample interpolation on the whole sample. During the course of carrying out 1/2 interpolation on the whole sample, firstly, a 4-order filter F1 is used for carrying out the interpolation wave filtration on the whole sample in the horizontal direction. Then, the 4-order filter F1 is used for carrying out the interpolation wave filtration in the vertical direction, and thus, the results of the 1/2 interpolation on the original sample value are obtained. During the course of carrying out 1/4 interpolation on the 1/2 interpolation results, firstly, a 4-order filter F2 and an image filter F3 are used for carrying out interpolation wave filtration on the 1/2 interpolation results in the horizontal direction; simultaneously, the 4-order filter F2 and the image filter F3 are used for carrying out interpolation wave filtration in the vertical direction; finally, the 1/4 samples in the four special positions are individually processed. The method has the characteristics of high interpolative prediction accuracy, and low time complexity and space complexity, and the method can be widely applied to the design of a system for encoding and decoding videos and images.
Description
Technical field
The present invention relates to the technical field of video and image coding and decoding, relating to especially is a kind of minute sample interpolation filtering method, is to utilize between frame of video image based on the encoding and decoding technique of time prediction.
Background technology
In video coding, often image is divided into two kinds of I picture and inter frame images.Wherein, inter frame image need adopt motion estimation coding method, and this mainly is because the pixel of adjacent image piece has very big temporal correlation.The main thought of this coding method is to find the piece that mates most with the encoding block predicted value (prediction piece) as encoding block in reference picture, encodes.Encoding block and prediction piece matching degree are high more, and the efficient of coding is high more.In order to improve the matching degree of the two, just need to improve the precision of estimation.What MPEG1 adopted is whole sample precision, MPEG2 and H.263 employing be 1/2 sample precision, in MPEG4, adopted 1/4 sample precision, thereby can improve code efficiency.But 1/4 precision estimation need be carried out 1/4 interpolation to reference picture.
In digital image processing field, 1/4 interpolation realizes by digital filter.The principle of digital filter is to act on the reference sample value by one group of filter coefficient, is inserted on the relevant position obtaining the predicted value of result as unknown sample.
In MPEG4,1/4 interpolation process computation complexity height, desire is carried out interpolation to a whole sample need carry out 6 rank linear interpolation and bilinear interpolations to the whole sample of the 6x6 around this sample.As shown in Figure 1, an image block through 1/4 interpolation after the size become original 16 times.A whole sample in the original image piece becomes 16 samples after through 1/4 interpolation.In addition, from space complexity, desire the sample block of the capable n row of n of interpolation, need use the whole sample of (n+5) row (n+5) row.We are through finding to exist following problems to anatomizing of interpolation process:
(1) the original sample image is carried out in the process of 1/2 interpolation, need be to carrying out 6 rank filtering, computation complexity height from nearest 6 the sample application filters (1/32 ,-5/32,20/32,20/32 ,-5/32,1/32) of interpolated sample on level or the vertical direction.
(2) sample point that relates in the interpolation process is many.
(3) carry out 1/4 interpolation and only adopt two samples to carry out linear averaging, be difficult to guarantee the interpolative prediction precision.
At above problem, we have proposed 1/4 new interpolation method of a cover.The advantage of this method is:
(1) adopts 4 rank filter F
1Original sample is carried out level and vertical direction 1/2 filtering interpolation, under the condition that guarantees the interpolation accuracy, reduced the sample number of computational complexity and participation computing, desire the data block of the capable n row of n of interpolation, only need the whole sample of (n+3) row (n+3) row, can effectively solve the bottleneck problem of memory access in the image coding and decoding process.
(2) adopt 4 rank filter F
2And mirror filter F
3Result to 1/2 interpolation carries out level and vertical direction 1/4 filtering interpolation, does not increase under the condition of computational complexity in assurance, has improved the accuracy of interpolative prediction.
(3) 1/4th samples of specific position are done individual processing, solve the problem of interpolative prediction distortion as a result in the sample rotary movement effectively.
For example, when the unit of each interpolation is 8 row, 8 row, this method is compared with the interpolation method among the MPEG4: average PSNR value this method of MPEG4 is hanged down 0.03dB, and the MPEG4 amount of calculation is bigger by 11.6% than this method, and the data volume of MPEG4 memory access simultaneously Duos 39.7% than this method.
Summary of the invention
The object of the present invention is to provide a kind of minute sample interpolation filtering method.This method may further comprise the steps:
(1) adopts 4 rank filter F
1Original sample is carried out level and vertical direction 1/2 filtering interpolation;
(2) adopt 4 rank filter F
2And mirror filter F
3Result to 1/2 interpolation carries out level and vertical direction 1/4 filtering interpolation;
(3) 1/4 sample to specific position carries out individual processing;
It is characterized in that: to be in whole sample and should whole sample around 3 1/2 samples, these 3 1/2 samples, distance is should whole sample nearest, and can with square of this whole composition of sample, 1/4 sample of the square center that is constituted carries out individual processing.
Following content provided 1/2 sample of luminance component image and the interpolation process of 1/4 sample.
If the whole sample of institute's reference is outside reference picture in interpolation process, should replace with the nearest whole sample (edge or angle sample) of the whole sample of distance reference in this image, promptly allow the outer sample of motion vector points reference picture.
Description of drawings
Fig. 1 is data block 1/4 an interpolation process schematic diagram of the present invention.
Fig. 2 is whole sample of the present invention (capitalization), 1/2 and 1/4 sample position (lowercase).
Shown in Figure 1, an image block through 1/4 interpolation after the size become original 16 times.A whole sample in the original image piece becomes 16 samples after through 1/4 interpolation.
Interpolation process is referring to Fig. 2, and the shaded block with the capitalization mark among Fig. 2 is the whole sample position of reference picture, is 1/2 and 1/4 sample position of reference picture with the transparent block of lowercase mark.
In following process, Clip1 (w) be defined as max (0, min (255, w)), 4 rank filter F
1With (q0, q1, q2, q3) expression, 4 rank filter F
2Coefficient with (p0, p1, p2, p3) expression, F
2Mirror filter F
3Coefficient with (p3, p2, p1, p0) expression, with " * " expression multiplying, with "/" expression division arithmetic.(q0, q1, q2, value q3) can be got (1/8,5/8,5/8 ,-1/8), (3/32,19/32,19/32 ,-3/32), (1/16,9/16,9/16 ,-1/16) or meet x/2
n, y/2
n, y/2
n, x/2
n, " x+y=2 wherein
N-1, 0<n<21, and n, x, y are integer " and (x/2 that requires
n, y/2
n, y/2
n, x/2
n). (p0, p1, p2, value p3) can be got (3/32,30/32,6/32 ,-1/32) or be met " p0+p1+p2+p3=2
n, wherein, 0<n<21, and n, p0, p1, p2, p3 are integer " and (p0/2 that requires
n, p1/2
n, p2/2
n, p3/2
n), filter F
3Select (1/32,6/32,30/32 ,-3/32) for use.
The luma prediction value of 1/2 sample position is by 4 rank filter F
1Calculate.Adopt 4 rank filter F
1Original sample is carried out level and vertical direction 1/2 filtering interpolation.When carrying out horizontal direction 1/2 interpolation, filter F
1Act on 4 whole samples of the nearest horizontal direction of the current interpolated sample of distance; When carrying out 1/2 interpolation of vertical direction, 1/2 sample of insertion and whole sample same column, filter F
1Act on 4 whole samples of the nearest vertical direction of the current interpolated sample of distance; When carrying out 1/2 interpolation of vertical direction, 1/2 sample of insertion and whole sample different lines, filter F
1Act on the median of 1/2 sample that 4 nearest apart from this target interpolation point on vertical direction level 1/2 interpolation obtain.
The luma prediction value of 1/4 sample position is by 4 rank filter F
2And mirror filter F
3Calculate.
When inserting 1/4 sample of going together with whole sample, 4 rank filter F
2And mirror filter F
3Act on 4 nearest whole samples of the horizontal direction current interpolated sample of distance; When inserting 1/4 sample with whole sample same column, 4 rank filter F
2And mirror filter F
3Act on 4 nearest whole samples of the vertical direction current interpolated sample of distance; When inserting 1/4 fen sample with whole sample different rows and different lines, if carry out 1/4 interpolation of horizontal direction, 4 rank filter F
2And mirror filter F
3Act on 1/2 sample of the current interpolated sample of horizontal direction distance nearest 4 and whole sample same column; If carry out 1/4 interpolation of vertical direction, 4 rank filter F
2And mirror filter F
3Act on 4 nearest 1/2 samples of going together of the current interpolated sample of distance on the vertical direction with whole sample.
It is characterized in that: to be in whole sample and should whole sample around 3 1/2 samples, these 3 1/2 samples, distance is should whole sample nearest, and can with square of this whole composition of sample, 1/4 sample of the square center that is constituted carries out individual processing.Insertion is positioned at 1/4 sample that whole sample the right 1 is listed as following 1 row, and this 1/4 sample value adopts this whole sample value and 1/2 sample values that should put in order following 2 row of sample the right 2 row to do linear averaging; Insertion is positioned at 1/4 sample that the whole sample left side 1 is listed as following 1 row, and this 1/4 sample value adopts this whole sample value and 1/2 sample values that should put in order following 2 row of the sample left side 2 row to do linear averaging; Insertion is positioned at 1/4 sample that whole sample the right 1 lists limit 1 row, and this 1/4 sample value adopts this whole sample value and 1/2 sample value of being somebody's turn to do 2 row tops, whole sample the right, 2 row to do linear averaging; Insertion is positioned at 1/4 sample that the whole sample left side 1 lists limit 1 row, and this 1/4 sample value adopts this whole sample value and 1/2 sample value that should put in order 2 row tops, the sample left side, 2 row to do linear averaging.
If the whole sample of institute's reference is outside the reference sample set in interpolation process, should replace with the nearest whole sample of the whole sample of distance reference in this sample set.
1/2 sample and 1/4 sample interpolation process process are as follows respectively:
1. 1/2 sample value that is labeled as b is obtained by following:
● when carrying out horizontal direction 1/2 interpolation, filter F
1Act on 4 whole samples of the nearest horizontal direction of the current interpolated sample of distance.At first with 4 nearest on the horizontal direction whole sample value C, D, E, F utilize 4 rank filter F
1Carry out filtering, obtain its median b ', as follows:
b’=((q0)*C+(q1)*D+(q2)*E+(q3)*F+1/2),
● final predicted value b is calculated as follows and obtains:
b=Clip1(b’)。
2. 1/2 sample value that is labeled as h is obtained by following:
● when carrying out 1/2 interpolation of vertical direction, 1/2 sample of insertion and whole sample same column, filter F
1Act on 4 whole samples of the nearest vertical direction of the current interpolated sample of distance.At first with 4 nearest on the vertical direction whole sample value A, D, H, K utilize 4 rank filter F
1Carry out filtering, and obtain its median h ', as follows:
h’=((q0)*A+(q1)*D+(q2)*H+(q3)*K+1/2),
● final predicted value h is calculated as follows and obtains:
h=Clip1(h’)。
3. 1/2 sample value that is labeled as j is obtained by following:
● when carrying out 1/2 interpolation of vertical direction, 1/2 sample of insertion and whole sample different lines, filter F
1Act on the median of 1/2 sample that 4 nearest apart from this target interpolation point on vertical direction level 1/2 interpolation obtain.At first on level or vertical direction to the median bb of 4 1/2 nearest sample values, h, m, cc or aa, b, s, dd utilize 4 rank filter F
1Carry out filtering, and obtain its median j ', as follows:
j’=((q0)*bb+(q1)*h+(q2)*m+(q3)*cc+1/2),
Perhaps
j’=((q0)*aa+(q1)*b+(q2)*s+(q3)*dd+1/2)。
Wherein, be labeled as aa, the 1/2 sample median of dd and s can be by utilizing 4 rank filter F on the horizontal direction
1Carry out filtering and obtain (identical) with the process of asking b '; Be labeled as bb, the 1/2 sample median of cc and m can be by utilizing 4 rank filter F on the vertical direction
1Carry out filtering and obtain (identical) with the process of asking h '.
● final predicted value j is calculated as follows and obtains:
j=Clip1(j’)。
The value that adopts the Filtering Processing on horizontal direction or the vertical direction to obtain is identical.
4. 1/4 sample value that is labeled as a and c is obtained by following:
● when inserting 1/4 sample of going together with whole sample, 4 rank filter F
2And mirror filter F
3Act on 4 nearest whole samples of the horizontal direction current interpolated sample of distance.At first in the horizontal direction to being labeled as C, D, four sample values of E and F are utilized 4 rank filter F
2And mirror filter F
3Carry out filtering, and obtain its median a ' and c ', as follows:
a’=((p0)*C+(p1)*D+(p2)*E+(p3)*F+1/2),
c’=((p3)*C+(p2)*D+(p1)*E+(p0)*F+1/2),
● final predicted value a and c are calculated as follows and obtain:
a=Clip1(a’),
c=Clip1(c’)。
Wherein, C, D, E and F are whole sample values.
5. 1/4 sample value that is labeled as d and n is obtained by following:
● when inserting 1/4 sample with whole sample same column, 4 rank filter F
2And mirror filter F
3Act on 4 nearest whole samples of the vertical direction current interpolated sample of distance.At first in vertical direction to being labeled as A, D, four sample values of H and K are utilized 4 rank filter F
2And mirror filter F
3Carry out filtering, and obtain its median d ' and n ', as follows:
d’=((p0)*A+(p1)*D+(p2)*H+(p3)*K+1/2),
n’=((p3)*A+(p2)*D+(p1)*H+(p0)*K+1/2),
● final predicted value d and n are calculated as follows and obtain:
d=Clip1(d’),
n=Clip1(n’)。
Wherein, A, D, H and K are whole sample values.
6. 1/4 sample value that is labeled as i and k is obtained by following:
● when inserting 1/4 fen sample with whole sample different rows and different lines, if carry out 1/4 interpolation of horizontal direction, 4 rank filter F
2And mirror filter F
3Act on 1/2 sample of the current interpolated sample of horizontal direction distance nearest 4 and whole sample same column.At first in the horizontal direction to being labeled as bb, h, four sample values of m and cc are utilized 4 rank filter F
2And mirror filter F
3Carry out filtering, and obtain its median i ' and k ', as follows:
i’=((p0)*bb+(p1)*h+(p2)*m+(p3)*cc+1/2),
k’=((p3)*bb+(p2)*h+(p1)*m+(p0)*cc+1/2),
● final predicted value i and k are calculated as follows and obtain:
i=Clip1(i’),
k=Clip1(k’)。
Wherein, bb, h, m and cc are 1/2 sample values.
7. 1/4 sample value that is labeled as f and q is obtained by following:
● when inserting 1/4 fen sample with whole sample different rows and different lines, if carry out 1/4 interpolation of vertical direction, 4 rank filter F
2And mirror filter F
3Act on the vertical direction 4 nearest 1/2 samples of going together of the current interpolated sample of distance at first in vertical direction to being labeled as aa with whole sample, b, four sample values of s and dd are utilized 4 rank filter F
2And mirror filter F
3Carry out filtering, and obtain its median f ' and q ', as follows:
f’=((p0)*aa+(p1)*b+(p2)*s+(p3)*dd+1/2),
q’=((p3)*aa+(p2)*b+(p1)*s+(p0)*dd+1/2),
● final predicted value f and q are calculated as follows and obtain:
f=Clip1(f’),
q=Clip1(q’)。
Wherein, aa, b, s and dd are 1/2 sample values.
8. to obtain being labeled as e, g, 1/4 sample value of p and r, computational process is as follows:
It is characterized in that: to be in whole sample and should whole sample around 3 1/2 samples, these 3 1/2 samples, distance is should whole sample nearest, and can with square of this whole composition of sample, 1/4 sample of the square center that is constituted carries out individual processing.
Insertion is positioned at the 1/4 sample e that whole sample D the right 1 is listed as following 1 row, and this 1/4 sample value adopts this whole sample value D and the 1/2 sample value j that should put in order following 2 row of sample the right 2 row to do linear averaging: e=(D+j+1)/2;
Insertion is positioned at the 1/4 sample g that the whole sample E left side 1 is listed as following 1 row, and this 1/4 sample value adopts this whole sample value E and the 1/2 sample value j that should put in order following 2 row of the sample left side 2 row to do linear averaging: g=(E+j+1)/2;
Insertion is positioned at the 1/4 sample p that whole sample H the right 1 lists limit 1 row, and this 1/4 sample value adopts this whole sample value H and the 1/2 sample value j that is somebody's turn to do 2 row tops, whole sample the right, 2 row to do linear averaging: p=(H+j+1)/2;
Insertion is positioned at the 1/4 sample r that the whole sample I left side 1 lists limit 1 row, and this 1/4 sample value adopts this whole sample value I and the 1/2 sample value j that should put in order 2 row tops, the sample left side, 2 row to do linear averaging: r=(I+j+1)/2.
The characteristics of this method are the accuracy height of interpolative prediction, and time complexity and space complexity are low.This method can be widely used in the design of video and image coding and decoding system.
Claims (18)
1. branch sample interpolation filtering method, this method may further comprise the steps:
(1) adopts 4 rank filter F
1Original sample is carried out level and vertical direction 1/2 filtering interpolation;
(2) adopt 4 rank filter F
2And mirror filter F
3Result to 1/2 interpolation carries out level and vertical direction 1/4 filtering interpolation;
(3) 1/4 sample to specific position carries out individual processing;
It is characterized in that: to be in whole sample and should whole sample around 3 1/2 samples, these 3 1/2 samples, distance is should whole sample nearest, and can with square of this whole composition of sample, 1/4 sample of the square center that is constituted carries out individual processing.
2. in accordance with the method for claim 1, it is characterized in that: when carrying out horizontal direction 1/2 interpolation, filter F
1Act on 4 whole samples of the nearest horizontal direction of the current interpolated sample of distance.
3. in accordance with the method for claim 1, when carrying out 1/2 interpolation of vertical direction, 1/2 sample of insertion and whole sample same column, filter F
1Act on 4 whole samples of the nearest vertical direction of the current interpolated sample of distance.
4. in accordance with the method for claim 1, when carrying out 1/2 interpolation of vertical direction, 1/2 sample of insertion and whole sample different lines, filter F
1Act on the median of 1/2 sample that 4 nearest apart from this target interpolation point on vertical direction level 1/2 interpolation obtain.
5. in accordance with the method for claim 1, it is characterized in that: when inserting 1/4 sample of going together with whole sample, 4 rank filter F
2And mirror filter F
3Act on 4 nearest whole samples of the horizontal direction current interpolated sample of distance.
6. in accordance with the method for claim 1, it is characterized in that: when inserting 1/4 sample with whole sample same column, 4 rank filter F
2And mirror filter F
3Act on 4 nearest whole samples of the vertical direction current interpolated sample of distance.
7. in accordance with the method for claim 1, it is characterized in that: when inserting 1/4 fen sample with whole sample different rows and different lines, if carry out 1/4 interpolation of horizontal direction, 4 rank filter F
2And mirror filter F
3Act on 1/2 sample of the current interpolated sample of horizontal direction distance nearest 4 and whole sample same column; If carry out 1/4 interpolation of vertical direction, 4 rank filter F
2And mirror filter F
3Act on 4 nearest 1/2 samples of going together of the current interpolated sample of distance on the vertical direction with whole sample.
8. according to claim 1 or 3 described methods, it is characterized in that: filter F
1Select (x/2 for use
n, y/2
n, y/2
n, x/2
n), x+y=2 wherein
N-1, 0<n<21, and n, x, y are integer.
9. according to claim 1 or 3 described methods, it is characterized in that: filter F
1Select (1/8,5/8,5/8 ,-1/8) for use.
10. according to claim 1 or 3 described methods, it is characterized in that: filter F
1Select (3/32,19/32,19/32 ,-3/32) for use.
11., it is characterized in that: filter F according to claim 1 or 3 described methods
1Select (1/16,9/16,9/16 ,-1/16) for use.
12. it is characterized in that in accordance with the method for claim 1: filter F
2Select (p0/2 for use
n, p1/2
n, p2/2
n, p3/2
n) and filter F
3Select (p3/2 for use
n, p2/2
n, p1/2
n, p0/2
n) p0+p1+p2+p3=2 wherein
n, wherein, 0<n<21, and n, p0, p1, p2, p3 are integer.
13. it is characterized in that in accordance with the method for claim 1: filter F
2Select (3/32,30/32,6/32 ,-1/32) and filter F for use
3Select (1/32,6/32,30/32 ,-3/32) for use.
14. according to claim 1 or 4 described methods, it is characterized in that: insert 1/4 sample be positioned at following 1 row of whole sample the right 1 row, this 1/4 sample value adopts this whole sample value and 1/2 sample values that should following 2 row of whole sample the right 2 row to do linear averaging.
15. according to claim 1 or 4 described methods, it is characterized in that: insert 1/4 sample be positioned at following 1 row of the whole sample left side 1 row, this 1/4 sample value adopts this whole sample value and 1/2 sample values that should following 2 row of the whole sample left side 2 row to do linear averaging.
16. according to claim 1 or 4 described methods, it is characterized in that: insert to be positioned at and put in order 1/4 sample that sample the right 1 lists limit 1 row, this 1/4 sample value adopts this whole sample value and 1/2 sample values that should put in order 2 row tops, sample the right, 2 row to do linear averaging.
17. according to claim 1 or 4 described methods, it is characterized in that: insertion is positioned at 1/4 sample that the whole sample left side 1 lists limit 1 row, and this 1/4 sample value adopts this whole sample value and 1/2 sample value that should put in order 2 row tops, the sample left side, 2 row to do linear averaging.
18. in accordance with the method for claim 1, if the whole sample of institute's reference is outside the reference sample set in interpolation process, should replace with the nearest whole sample of the whole sample of distance reference in this sample set.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN 200310101370 CN1216498C (en) | 2003-10-20 | 2003-10-20 | Sub-sample interpolation filter method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN 200310101370 CN1216498C (en) | 2003-10-20 | 2003-10-20 | Sub-sample interpolation filter method |
Publications (2)
Publication Number | Publication Date |
---|---|
CN1529412A CN1529412A (en) | 2004-09-15 |
CN1216498C true CN1216498C (en) | 2005-08-24 |
Family
ID=34304195
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN 200310101370 Expired - Lifetime CN1216498C (en) | 2003-10-20 | 2003-10-20 | Sub-sample interpolation filter method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN1216498C (en) |
Families Citing this family (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP4420294B2 (en) * | 2005-06-28 | 2010-02-24 | エルジー ディスプレイ カンパニー リミテッド | Midian filtering method and apparatus |
CN1794821A (en) * | 2006-01-11 | 2006-06-28 | 浙江大学 | Method and device of interpolation in grading video compression |
CN101753799B (en) * | 2008-12-16 | 2012-07-11 | 无锡华润矽科微电子有限公司 | Median filtering method and median filtering circuit using the same |
-
2003
- 2003-10-20 CN CN 200310101370 patent/CN1216498C/en not_active Expired - Lifetime
Also Published As
Publication number | Publication date |
---|---|
CN1529412A (en) | 2004-09-15 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN1812592A (en) | Method and apparatus for processing image data of a color filter array | |
CN1287598C (en) | Filtering method and apparatus for removing blocking artifacts and/or ringing noise | |
CN1668108A (en) | Resolution changing method and apparatus | |
CN1708134A (en) | Method and apparatus for estimating motion | |
CN1246765A (en) | Method and system for digital signal interpolation | |
CN1926875A (en) | Motion compensation method | |
CN1607836A (en) | De-blocking filter processing apparatus and de-blocking filter processing method | |
CN1256849C (en) | Method and apparatus for 1/4 pixel precision interpolation | |
CN101068364A (en) | Video encoder and graph processing unit | |
CN101080930A (en) | Motion vector calculation device and motion vector calculation method | |
CN1960495A (en) | Picture coding device, method, and program thereof | |
CN1767644A (en) | Non-integer pixel sharing for video encoding | |
CN1310513C (en) | Image processing device and method, video display device, and recorded information reproduction device | |
CN1951121A (en) | Block noise reducing device and image display | |
CN1722842A (en) | The filtering method of audio-visual codec and filter apparatus | |
CN1863319A (en) | Fractional pixel interpolation method and encoding and decoding apparatus | |
CN1498501A (en) | Method for compressing and decompressing video data | |
CN100342710C (en) | Structure method for enhancing image | |
CN1791221A (en) | Method for realizing video code anti-shake based on dynamic image and coder | |
CN1216498C (en) | Sub-sample interpolation filter method | |
CN1529509A (en) | Video image sub-picture-element interpolation method and device | |
WO2019196941A1 (en) | Adaptive implicit transform setting | |
CN1723712A (en) | Joint resolution or sharpness enhancement and artifact reduction for coded digital video | |
CN1633161A (en) | A method for realizing integral multiple amplification of image | |
CN1492689A (en) | Separate sample interpolation filter method |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
C14 | Grant of patent or utility model | ||
GR01 | Patent grant | ||
CX01 | Expiry of patent term | ||
CX01 | Expiry of patent term |
Granted publication date: 20050824 |