CN103647919B - Edge-sensitive intra-field de-interlacing method - Google Patents
Edge-sensitive intra-field de-interlacing method Download PDFInfo
- Publication number
- CN103647919B CN103647919B CN201310724193.0A CN201310724193A CN103647919B CN 103647919 B CN103647919 B CN 103647919B CN 201310724193 A CN201310724193 A CN 201310724193A CN 103647919 B CN103647919 B CN 103647919B
- Authority
- CN
- China
- Prior art keywords
- pixel
- row vector
- interpolation
- gradient
- point
- 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 - Fee Related
Links
Abstract
The invention relates to an edge-sensitive intra-field de-interlacing method. The edge-sensitive intra-field de-interlacing method comprises the following steps of 1 performing interpolation on all of pixel points to be interpolated by means of a method for obtaining a mean value of every two adjacent pixel points in the vertical direction; 2 utilizing a gradient operator to calculate the gradients and gradient magnitudes of each pixel point to be interpolated in the horizontal direction and the vertical direction according to known pixel points; 3 judging regions where pixel points to be interpolated are located according to the gradient magnitudes; 4 selecting a plurality of pixel points in a complicated region to form three row vectors; 5 selecting row vectors for follow-up calculation according to the corresponding relation of the gradients in the horizontal direction and the vertical direction; 6 calculating a difference value of each pixel point in the selected row vectors and an existing pixel point to obtain a pixel point with a minimum difference value absolute value, namely a most similar pixel point; 7 performing interpolation on the pixel point to be interpolated corresponding to the position of the most similar pixel point. Compared with the prior art, the edge-sensitive intra-field de-interlacing method has the advantages of being low in computation complexity, high in accuracy and the like.
Description
Technical field
The present invention relates to the interlace-removing method of a kind of image or video, especially relate to go in a kind of field of edge sensitive every
Row method.
Background technology
In the television broadcasting system of the analog systems such as ntsc, pal and secam, horizontal-interlace technique has had become as mark
Standard, and digital TV in high resolution broadcast system now also remains this scan format.Because it can not only save transmission of video
Bandwidth, and existing transmission channel can be maintained.However, with novel flat-plates such as liquid crystal display, plasma displays
The development of display, the video of progressive scan becomes mainstream demand gradually.Because all emerging display are almost required for
Progressive inputs.And, interleaved video occurs that the bad vision such as interline flicker, jump and edge sawtooth is imitated
Really.In order to overcome these problems, the deinterlacing technique that interlaced scanning video can be converted into progressive scanned video arises at the historic moment.
Deinterlacing technique is broadly divided into frame in de interlacing and two kinds of interframe de interlacing.In general, interframe deinterlacing technique
De interlacing effect better than frame in deinterlacing technique is obtained due to the Pixel Information using multiple frame of video, but this also increases
The cost that hardware is realized.And frame in deinterlacing technique only need store several rows Pixel Information just can obtain go well every
Row effect, this not only greatlys save hardware resource, and can also be used as the key component of interframe de interlacing algorithm.Most common
Frame in de interlacing algorithm be that two points up and down treating interpolating pixel point are averaged, although this method is simple, meeting
Bring serious sawtooth effect at edge.Frame in interlace-removing method based on rim detection usually can cause because of missing inspection and picking up by mistake
The loss of image detail is it is impossible to obtain gratifying de interlacing effect.
Content of the invention
The purpose of the present invention be exactly provide to overcome the defect that above-mentioned prior art exists a kind of computation complexity low,
The in-field deinterlacing method of the edge sensitive of high precision.
The purpose of the present invention can be achieved through the following technical solutions:
A kind of in-field deinterlacing method of edge sensitive, including following methods:
1) all interpolation pixels being carried out with vertical direction asks the method for adjacent two pixel averages to enter row interpolation；
2) according to known pixels point, calculate each interpolation pixel horizontally and vertically using gradient operator
On gradient gx, gy and gradient magnitude magg；
3) region according to residing for gradient magnitude judges interpolation pixel: if magg > th1, it is in complex region, if
Magg≤th1, then be in flat site, and wherein, th1 is Grads threshold；
4) choose some pixels in complex region and form three row vectors, including up row vector, down row vector and
Side row vector；
5) row needed for subsequent calculations being chosen according to the corresponding relation of gradient gx, gy horizontally and vertically going up to
Amount: if | gx | is ＜=a | gy |, select up row vector and down row vector to carry out subsequent calculations, execution step 6) simultaneously；If |
Gx | ＞ a | gy |, then select side row vector to carry out subsequent calculations, execution step 6), wherein, a is regulation coefficient；
6) pixel in row vector selected by calculating and the difference of current pixel point, obtain the minimum pixel of absolute difference
Point, as most like pixel；
7) according to step 6) position of most like pixel that obtains enters row interpolation to corresponding interpolation pixel.
Described gradient operator includes horizontal direction gradient operator and vertical gradient operator.
The span of described Grads threshold th1 is 500～1500.
Described up row vector is by three pixels on the right of three, the next line left side pixel of current pixel point and lastrow
Point composition；Described down row vector is by three pictures on the right of three, the lastrow left side pixel of current pixel point and next line
Vegetarian refreshments forms；Described side row vector is by five pixels of left side string of current pixel point and five pixels of the right string
Composition.
The span of described regulation coefficient a is 1～8.
Described step 6) in, if selected row vector is up row vector and down row vector, proceed as follows:
A) judge whether the minimum pixel of the absolute difference obtaining in up row vector meets minup < th2 simultaneously and be somebody's turn to do
Pixel corresponding gx and gy jack per line, if so, then using this pixel as most like pixel, if it is not, then execution step b), its
In, minup is the minima of the absolute difference in up row vector, and th2 is given threshold；
B) judge whether the minimum pixel of the absolute difference obtaining in down row vector meets mindown < th2 simultaneously
And this pixel corresponding gx and gy contrary sign, if so, then using this pixel as most like pixel, if it is not, processing next
Pixel, wherein, mindown is the minima of the absolute difference in down row vector.
Described step 7) in, most like pixel has five types, and each type of pixel is adopted with different inserting
Value processing method:
First type corresponding interpolation pixel is be 1 with current pixel point horizontal range four nearest to be inserted
Value pixel, corresponding interpolation process method is to take 4 points of averages according to edge direction；
Second type corresponding interpolation pixel is be 0 with current pixel point horizontal range two nearest to be inserted
Value pixel, does not carry out interpolation processing；
The third type corresponding interpolation pixel is be 1 with current pixel point horizontal range four nearest to be inserted
Value pixel, corresponding interpolation process method is to take 2 points of averages according to edge direction；
The corresponding interpolation pixel of 4th type is be 2 with current pixel point horizontal range four nearest to be inserted
Value pixel, corresponding interpolation process method is to take 2 points of averages according to edge direction；
Remaining is the 5th type, does not carry out interpolation processing.
Compared with prior art, the invention has the advantages that
1) obtain preferable de interlacing effect on the basis of taking relatively low hardware resource, maintain video figure well
The edge of picture and details.And in order to reduce computation complexity, flat site is asked using vertical direction with the algorithm of 2 points of averages.
2) in order to improve the precision of gradient calculation, reduce effect of noise, be specifically designed gradient operator it is achieved that more
Accurately edge judges in advance so that the selection to interpolation point more has specific aim.
3) the closest pixel different to five kinds, using different interpolation methods, serves to the edge of video image
Preferably recovery effects.
Brief description
Fig. 1 is the schematic flow sheet of the present invention；
Fig. 2 is the gradient operator of the present invention；
(2a) for the gradient operator of horizontal direction, (2b) is the gradient operator of vertical direction；
Fig. 3 is the schematic diagram in region residing for current pixel point in the inventive method；
Fig. 4 is three row vector schematic diagrams of the present invention；
(4a) it is up row vector, (4b) is down row vector, (4c) is side row vector；
Fig. 5 is interpolation method schematic diagram of the present invention.
Specific embodiment
The present invention is described in detail with specific embodiment below in conjunction with the accompanying drawings.
As shown in figure 1, a kind of in-field deinterlacing method of edge sensitive, comprise the following steps:
In step s01, according to specially designed gradient operator calculate the gradient horizontally and vertically going up with
And the amplitude of gradient.The gradient operator that this method adopts includes the gradient operator of horizontal direction and the gradient operator of vertical direction,
The weight distribution of each gradient operator is as shown in Figure 2.Calculate the horizontal gradient gx of each interpolation pixel using this two operators
With vertical gradient gy.Its computing formula is as follows:
Gx=f (i-4, j-2)+4f (i-4, j-1) -4f (i-4, j+1)-f (i-4, j+2)+2f (i-2, j-2)+5f (i-2,
j-1)-5f(i-2,j+1)-2f(i-2,j+2)+6f(i,j-2)+6f(i,j-1)-6f(i,j+1)-6f(i,j+2)+2f(i+2,
j-2)+5f(i+2,j-1)-5f(i+2,j+1)-2f(i+2,j+2)+f(i+4,j-2)+4f(i+4,j-1)-4f(i+4,j+1)-f
(i+4,j+2)；
Gy=f (i-4, j-2)+2f (i-4, j-1)+6f (i-4, j)+2f (i-4, j+1)+f (i-4, j+2)+4f (i-2, j-
2)+5f(i-2,j-1)+6f(i-2,j)+5f(i-2,j+1)+4f(i-2,j+2)-4f(i+2,j-2)-5f(i+2,j-1)-6f(i
+2,j)-5f(i+2,j+1)-4f(i+2,j+2)-f(i+4,j-2)-2f(i+4,j-1)-6f(i+4,j)-2f(i+4,j+1)-f
(i+4,j+2).
The computing formula of each interpolation pixel gradient magnitude magg is as follows:
Flat site and the complex region of image or video frame in step s02, is gone out according to the magnitude resolution of gradient.
It is illustrated in figure 3 the schematic diagram in region residing for current pixel point.Pixel, the white round dot known to the representative of black round dot of in figure
Represent interpolation pixel, Lycoperdon polymorphum Vitt round dot represents current pixel point.Due to calculating residing for the pixel that gradient adopts in step s01
Scope is larger it is believed that the interpolation pixel showing in Fig. 3 is in the same area with current pixel point.
Grads threshold th1 is set gradient to be analyzed judge, the span of th1 is 500～1500.Work as magg >
During th1, pixel is in complex region, as magg≤th1, when vegetarian refreshments is in flat site.Do not carrying out region division
Under original state, acquiescence all pixels point, all in flat site, is treated interpolating pixel and is sought 2 points of averages using vertical direction
Algorithm.After carrying out region division, the interpolation pixel being in complex region is updated, and to being in flat site
Pixel is not further processed.
In step s03, in complex region, form three row vectors according to specific position relationship selected pixels point,
They are up row vector, down row vector and side row vector respectively.As shown in figure 4, up row vector (figure (4a)) is by current
Three pixel compositions, i.e. up={ f (i+2, j-4), f on the right of three, the next line left side of pixel pixel and lastrow
(i+2,j-2),f(i+2,j),f(i-2,j),f(i-2,j+2),f(i-2,j+4)}.Down row vector (figure (4b)) is by current picture
Three pixel compositions, i.e. down={ f (i-2, j-4), f on the right of three, the lastrow left side of vegetarian refreshments pixel and next line
(i-2,j-2),f(i-2,j),f(i+2,j),f(i+2,j+2),f(i+2,j+4)}.Side row vector (figure (4c)) is by current picture
Five pixels of the left side string of vegetarian refreshments and five pixels of the right string composition, that is, side=f (i-4, j-2), f (i-2,
j-2),f(i,j-2),f(i+2,j-2),f(i+4,j-2),f(i-4,j+2),f(i-2,j+2),f(i,j+2),f(i+2,j+
2),f(i+4,j+2)}.
In step s04, the corresponding relation according to gradient gx, gy horizontally and vertically going up chooses subsequent calculations
Required row vector.As | gx | ＜=a | gy |, up row vector and down row vector is selected to carry out subsequent calculations, execution step
s06.As | gx | ＞ a | gy |, side row vector is selected to carry out subsequent calculations, execution step s05.The order of magnitude of gx and gy
React the angular dimension at edge.For example as | gx |=| gy |, can roughly think that edge is in 45 degree of angles or 135 degree
Angle.When edge angle closer to vertical when, select side row vector carry out subsequent calculations.When edge angle is closer to level,
Up row vector and down row vector is selected to carry out subsequent calculations.
Parameter a is a regulation coefficient, can adjust the proportionate relationship of | gx | and | gy |, so as to the selection of row vector more
Added with effect, its span is 1～8.
In step s05, calculate the difference of the pixel in side row vector and current pixel point, difference is taken absolute value
Sequence obtains minima and the position of corresponding pixel afterwards, obtains most like pixel, execution step s09.
In step s06, calculate the difference of the pixel in up row vector and down row vector and current pixel point respectively,
After difference is taken absolute value, sequence obtains minima and the position of corresponding pixel, and the two is done selects further.
In step s07, if up row vector asks difference to sort minima minup obtaining after taking absolute value less than threshold value
Th2, and gx and gy jack per line (be all on the occasion of or be all negative value), then just by most like pixel corresponding in up row vector
For entering row interpolation in step s09, otherwise execution step s08.Wherein minup is the minimum of the absolute difference in up row vector
Value, th2 is given threshold.
In step s08, if down row vector asks difference to sort minima mindown obtaining after taking absolute value less than threshold
Value th2, and gx and gy contrary sign (positive is negative), then just most like pixel corresponding in down row vector is used for step
Row interpolation is entered, wherein mindown is the minima of the absolute difference in down row vector in s09.
If up and down row vector is all unsatisfactory for condition, then just do not treat to being in the same area with current pixel point
Interpolating pixel point is further processed, and processes next pixel.
In step s09, according to the calculated and immediate pixel of current pixel point pixel value come to corresponding positions
The pixel put enters row interpolation, thus obtaining preferable de interlacing effect.As shown in figure 5, having the most like of five types
Pixel, every kind of pixel may correspond to different interpolation pixels, so needing to be treated differently.
For the first type, have four pixels and belong to this type, they are respectively, f (i-4, j+2), f (i+
4, j+2), f (i-4, j-2) and f (i+4, j-2), corresponding interpolation pixel is be 1 with current pixel point horizontal range four
Individual nearest interpolation pixel.Here just taking current pixel point upper right comer region as a example illustrate, there is class in other regions
As processing mode.If most like pixel is f (i-4, j+2), then the direction at edge is exactly current pixel point and f (i-
4, j+2) direction of line, and then just select interpolation pixel f (i-1, j+1) to enter row interpolation.Interpolation method is according to side
Edge direction takes 4 points of average: f (i-1, j+1)=(f (i-2, j+1)+f (i-2, j+2)+f (i, j)+f (i, j+1))/4.Due to treating
Known pixels point is not had on the edge residing for interpolating pixel point f (i-1, j+1), therefore chooses here from nearest four in this edge
Known pixels point is entering row interpolation, and also makes the edge direction after interpolation with current pixel using the average of this four pixels
Point is reached unanimity with the direction of f (i-4, j+2) line.
For second type, have two pixels and belong to this type, they are respectively, f (i-2, j) and f (i+2,
j).If most like pixel be f (i-2, j), corresponding interpolation pixel is be 0 with current pixel point horizontal range two
Individual nearest interpolation pixel.The direction at edge is just vertically oriented, and then (i-1 j) comes just to select interpolation pixel f
Enter row interpolation.Interpolation method is f (i-1, j)=(f (i, j)+f (i-2, j))/2.But, because interpolation pixel is all given tacit consent to
Interpolation is carried out using the algorithm that vertical direction seeks 2 points of averages, so interpolation need not be repeated here.If most like pixel
Be f (i+2, j), then be not also further processed.
For the third type, have four pixels and belong to this type, they are respectively, f (i-2, j+2), f (i+
2, j+2), f (i-2, j-2) and f (i+2, j-2), corresponding interpolation pixel is be 1 with current pixel point horizontal range four
Individual nearest interpolation pixel.Here just taking current pixel point upper right comer region as a example illustrate, there is class in other regions
As processing mode.If most like pixel is f (i-2, j+2), then the direction at edge is exactly current pixel point and f (i-
2, j+2) direction of line, and then just select interpolation pixel f (i-1, j+1) to enter row interpolation.Interpolation method is according to side
Edge direction takes 2 points of average: f (i-1, j+1)=(f (i, j)+f (i-2, j+2))/2.
For the 4th type, have four pixels and belong to this type, they are respectively, f (i-2, j+4), f (i+
2, j+4), f (i-2, j-4) and f (i+2, j-4), corresponding interpolation pixel is be 2 with current pixel point horizontal range four
Individual nearest interpolation pixel.Here just taking current pixel point upper right comer region as a example illustrate, there is class in other regions
As processing mode.If most like pixel is f (i-2, j+4), then the direction at edge is exactly current pixel point and f (i-
2, j+4) direction of line, and then just select interpolation pixel f (i-1, j+2) to enter row interpolation.Interpolation method is according to side
Edge direction takes 2 points of average: f (i-1, j+1)=(f (i, j)+f (i-2, j+4))/2.
For the 5th type, have two pixels and belong to this type, they are f (i, j+2) and f (i, j- respectively
2).If most like pixel is f (i, j+2), then the direction at edge is just horizontally oriented.But, due to interpolation picture
The pixel that vegetarian refreshments is in same a line is not known point, if all with unknown point come interpolation, error can be very big, so here not
Do interpolation processing.If most like pixel is f (i, j-2), then be not also further processed.
Above-mentioned specific embodiment is exemplary, be in order to preferably make skilled artisans appreciate that this
Invent it is impossible to be not understood as the present invention is included with the restriction of scope, as long as in the essential scope of the present invention, to above-mentioned enforcement
Any equivalent change or modification that example is made, each fall within the range of the claim of the present invention.
Claims (5)
1. a kind of in-field deinterlacing method of edge sensitive is it is characterised in that comprise the following steps:
1) all interpolation pixels being carried out with vertical direction asks the method for adjacent two pixel averages to enter row interpolation；
2) according to known pixels point, calculate horizontally and vertically going up of each interpolation pixel using gradient operator
Gradient gx, gy and gradient magnitude magg；
3) region according to residing for gradient magnitude judges interpolation pixel: if magg > th1, it is in complex region, if magg
≤ th1, then be in flat site, and wherein, th1 is Grads threshold；
4) choose some pixels in complex region and form three row vectors, including up row vector, down row vector and side
Row vector, described up row vector is by three pixels on the right of three, the next line left side pixel of current pixel point and lastrow
Composition；Described down row vector is by three pixels on the right of three, the lastrow left side pixel of current pixel point and next line
Point composition；Described side row vector is by five pixels of left side string of current pixel point and five pixel groups of the right string
Become；
5) row vector needed for the corresponding relation according to gradient gx, gy horizontally and vertically going up chooses subsequent calculations: if |
Gx | ＜=a | gy |, then select up row vector and down row vector to carry out subsequent calculations, execution step 6) simultaneously；If | gx | is ＞ a |
Gy |, then select side row vector to carry out subsequent calculations, execution step 6), wherein, a is regulation coefficient；
6) pixel in row vector selected by calculating and the difference of current pixel point, obtain the minimum pixel of absolute difference,
It is most like pixel；
7) according to step 6) position of most like pixel that obtains enters row interpolation to corresponding interpolation pixel.
2. a kind of in-field deinterlacing method of edge sensitive according to claim 1 is it is characterised in that described gradient is calculated
Attached bag includes horizontal direction gradient operator and vertical gradient operator.
3. a kind of in-field deinterlacing method of edge sensitive according to claim 1 is it is characterised in that described gradient threshold
The span of value th1 is 500～1500.
4. a kind of in-field deinterlacing method of edge sensitive according to claim 1 is it is characterised in that described adjustment system
The span of number a is 1～8.
5. a kind of in-field deinterlacing method of edge sensitive according to claim 1 is it is characterised in that described step 6)
In, if selected row vector is up row vector and down row vector, proceed as follows:
A) judge whether the minimum pixel of the absolute difference obtaining in up row vector meets minup < th2 and this pixel simultaneously
The corresponding gx and gy jack per line of point, if so, then using this pixel as most like pixel, if it is not, then execution step b), wherein,
Minup is the minima of the absolute difference in up row vector, and th2 is given threshold；
B) judge whether the minimum pixel of the absolute difference obtaining in down row vector meets mindown < th2 simultaneously and be somebody's turn to do
Pixel corresponding gx and gy contrary sign, if so, then using this pixel as most like pixel, if it is not, processing next pixel
Point, wherein, mindown is the minima of the absolute difference in down row vector.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310724193.0A CN103647919B (en) | 2013-12-24 | 2013-12-24 | Edge-sensitive intra-field de-interlacing method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310724193.0A CN103647919B (en) | 2013-12-24 | 2013-12-24 | Edge-sensitive intra-field de-interlacing method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103647919A CN103647919A (en) | 2014-03-19 |
| CN103647919B true CN103647919B (en) | 2017-01-18 |
Family
ID=50253065
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201310724193.0A Expired - Fee Related CN103647919B (en) | 2013-12-24 | 2013-12-24 | Edge-sensitive intra-field de-interlacing method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103647919B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109784213B (en) * | 2018-12-27 | 2022-10-11 | 合肥工业大学 | Low-cost non-maximum suppression circuit structure |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1477869A (en) * | 2002-07-26 | 2004-02-25 | ���ǵ�����ʽ���� | Interlacing-removing device and method |
| CN101309376A (en) * | 2008-06-13 | 2008-11-19 | 北京中星微电子有限公司 | Method and device for eliminating alternate line |
| CN101600061A (en) * | 2009-07-09 | 2009-12-09 | 杭州士兰微电子股份有限公司 | De-interlaced method of video motion-adaptive and device |
| US8081256B2 (en) * | 2007-03-20 | 2011-12-20 | Samsung Electronics Co., Ltd. | Method and system for edge directed deinterlacing in video image processing |
| CN102395019A (en) * | 2011-11-22 | 2012-03-28 | 华亚微电子（上海）有限公司 | Deinterlacing method |
-
2013
- 2013-12-24 CN CN201310724193.0A patent/CN103647919B/en not_active Expired - Fee Related
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1477869A (en) * | 2002-07-26 | 2004-02-25 | ���ǵ�����ʽ���� | Interlacing-removing device and method |
| US8081256B2 (en) * | 2007-03-20 | 2011-12-20 | Samsung Electronics Co., Ltd. | Method and system for edge directed deinterlacing in video image processing |
| CN101309376A (en) * | 2008-06-13 | 2008-11-19 | 北京中星微电子有限公司 | Method and device for eliminating alternate line |
| CN101600061A (en) * | 2009-07-09 | 2009-12-09 | 杭州士兰微电子股份有限公司 | De-interlaced method of video motion-adaptive and device |
| CN102395019A (en) * | 2011-11-22 | 2012-03-28 | 华亚微电子（上海）有限公司 | Deinterlacing method |
Non-Patent Citations (2)
| Title |
|---|
| "一种新型运动自适应去隔行算法";刘政林, 刘菊, 邹雪城, 陈毅成;《微处理机》;20081231;第29卷(第6期);全文 * |
| "加权边沿自适应的场内插值去隔行方法";马斌，丁勇，周升，严晓浪;《计算机应用研究》;20100831;第27卷(第8期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103647919A (en) | 2014-03-19 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101600061B (en) | Video motion-adaptive de-interlacing method and device therefor | |
| US6262773B1 (en) | System for conversion of interlaced video to progressive video using edge correlation | |
| US7423691B2 (en) | Method of low latency interlace to progressive video format conversion | |
| EP0830018A2 (en) | Method and system for motion detection in a video image | |
| JP3583684B2 (en) | Image defect detection apparatus and image defect detection method | |
| US8331689B2 (en) | Detecting a border region in an image | |
| CN102025960A (en) | Motion compensation de-interlacing method based on adaptive interpolation | |
| US9786038B2 (en) | Method for processing an image sequence having consecutive video images in order to improve the spatial resolution | |
| CN101699856B (en) | De-interlacing method with self-adapting motion | |
| US7460734B2 (en) | Method and apparatus of adaptive interpolation based on edge detection | |
| CN103647919B (en) | Edge-sensitive intra-field de-interlacing method | |
| JP2009207137A (en) | Method and system of processing video signal, and computer-readable storage medium | |
| CN102170549A (en) | Edge correlation image intra-field de-interlacing algorithm of edge pre-judgment | |
| CN101309376B (en) | Method and device for eliminating alternate line | |
| KR101158847B1 (en) | Deinterlacing apparatus and method using edge map | |
| CN106303338B (en) | A kind of in-field deinterlacing method based on the multi-direction interpolation of bilateral filtering | |
| KR100931110B1 (en) | Deinterlacing apparatus and method using fuzzy rule-based edge recovery algorithm | |
| US20080111916A1 (en) | Image de-interlacing method | |
| CN102281459A (en) | Motion vector display circuit and motion vector display method | |
| US6992727B2 (en) | Method for detecting video frame types with adaptive thresholds | |
| CN105898179A (en) | Deinterlace method and device for interlaced video | |
| CN107517356B (en) | Deinterlacing image processing method and device | |
| US20040207752A1 (en) | Method for detecting video frame types with median filtering | |
| KR960012490B1 (en) | Video format converting apparatus and method for hdtv | |
| KR100628190B1 (en) | Converting Method of Image Data's Color Format |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | 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 | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20170118 Termination date: 20191224 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |