CN100403769C - Image crisperding method during composing process - Google Patents
Image crisperding method during composing process Download PDFInfo
- Publication number
- CN100403769C CN100403769C CNB2006100786382A CN200610078638A CN100403769C CN 100403769 C CN100403769 C CN 100403769C CN B2006100786382 A CNB2006100786382 A CN B2006100786382A CN 200610078638 A CN200610078638 A CN 200610078638A CN 100403769 C CN100403769 C CN 100403769C
- Authority
- CN
- China
- Prior art keywords
- image
- crisperding
- typeset
- target
- algorithm
- 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
Images
Landscapes
- Image Processing (AREA)
Abstract
The present invention relates to an image crispening method during a composing process, which belongs to the technical field of image processing. Although existing image crispening technique is mature, crispening processing for images is rarely utilized in the composing process. Moreover, an existing image crispening algorithm can not be applied to the composing process directly. The method provided by the present invention has the technical scheme that an image in composition process is converted into a gray level image firstly; then, the gray level image is converted into a binary image; when the binary image is carried out the crispening processing, image element matrix of the image is magnified for integral times according to an equality ratio and a Freeman chain code following algorithm is adopted for obtaining a close Bezier curve of an object outline. The crispening processing of the image can be conveniently, high-effectively accomplished in the compositing process via the adoption of the method of the present invention. Moreover, the method is suitable for each type of image and is convenient for subsequent processing of the image. Simultaneously, users can adjust threshold values and tolerance to obtain different crispening results according to different requirements.
Description
Technical field
The invention belongs to technical field of image processing, be specifically related to a kind of method of in process of typeset, carrying out the image crisperding.
Background technology
Along with the maturation of image processing techniques, many image processing softwares need that all image is carried out crisperding and handle to obtain the content that the user wants when carrying out image processing.So-called image crisperding is meant target (being called image segmentation again) in the image is cut apart and is kept to the background in the image and target, obtains the Bezier curve representation that objective contour seals on this basis.Image Segmentation Technology is an important content in the image processing field, through years of development, several different methods occurred and obtained good application in various fields at present.
Present image partition method mainly contains following several method: based on gradient, based on the zone, based on the dividing method of Specific Theory Tools.
(1) based on the dividing method of gradient:
Relatively Chang Yong gradient operator has Sobel, Roberts, Canny operator etc., and the gray scale edge is the discontinuous result of gray value, and this discontinuity often can be utilized to differentiate and detects easily.These operators have reflected the single order or the second dervative of gray value, thereby directly can obtain edge of image.But this method is handling of a pixel of a pixel, does not consider the relation between pixel, and the edge that obtains is discontinuous often, also needs to do some more complicated work in order to obtain continuous edge.
(2) based on the dividing method in zone:
These class methods comprise the threshold transformation method based on pixel value, coordinate position, based on histogram method and based on dividing methods such as the method for space clustering and region growing, division merging.These class methods are fairly simple, target and background can be divided into different zones, utilize various path trace algorithms to obtain the profile of target then.But, in order to cut apart target more accurately, needing the relation between considered pixel, the complexity that this has just caused algorithm causes the reduction of efficient.
(3) based on the dividing method of Specific Theory Tools:
Because the image source is very extensive, the content that comprises in the image is very abundant, and simple dividing method does not satisfy the requirement of each side.For this reason, scholars have been merged the knowledge in each field, have proposed as based on dividing methods such as mathematical morphology, pattern recognition, neural net, information theories.These methods can access effect preferably, but because algorithm is very complicated, say it also is minimum on efficient, and every kind of method only has good result for a certain class or a few class image, and are unsatisfactory for other types of image effect.
With image object and background area separately after, in order to obtain the sealing Bezier curve representation of objective contour, need carry out profile to the target area and follow the tracks of.Profile track algorithm commonly used at present has deformable template track algorithm, Snake algorithm and Freeman chain code following algorithm.The deformable template track algorithm depends on the priori of shape of image object and very sensitive to picture noise; The Snake algorithm is followed the tracks of based on the energy minimization principle, and amount of calculation is big, efficient is lower; Freeman chain code following algorithm adopts 3 * 3 templates, and track algorithm is simple, but is easy to obtain wrong profile information for complicated shape.
Freeman chain code following algorithm detailed process is as follows:
(a) determine initial tracking direction: according to the definition of F chain code, starting point is sought such point successively from 0-＞7 directions, this F chain code direction with respect to starting point is dir, this point is impact point, and the point on (dir+1) %8 direction of starting point is a background dot, and then this is the next profile search point of starting point;
(b) coordinate figure that will put and dir deposit the Freeman structure in, insert in the chained list;
(c) according to current point (xcur, ycur) and direction dir find next point (xnext, ynext) and to make it be current point;
(d) if current point (xcur, ycur)=(xstart ystart), then finishes route searching, forwards step (e) to, otherwise from the dir direction:
I is if the point that dir points to is a background dot, and then dir=(dir+7) %8 carries out i once more, and the point that points to up to dir is an impact point, and then dir is the chain code of current point, changes step (b);
Ii then makes dir=(dir+1) %8 if the point that dir points to is an impact point, carries out ii again, and the point that points to up to dir is a background dot, and the chain code of then current point is dir=(dir+7) %8, changes step (b);
(e) behind the profile that obtains a target area, the tracking initiation point in the next zone of search is proceeded the Freeman chain code following again, finishes up to all target areas are followed the tracks of; At last fit to the Bezier curve according to the value in each regional Freeman chain code.
Although in image processing software, image is carried out crisperding handled very maturation, this technology is used seldom in process of typeset.Along with the develop rapidly of composing industry, the object that need set type becomes increasingly complex, and particularly a lot of images is needed to carry out crisperding too and handles.But, the technology of also not handling in the present software for composing at the image crisperding.And; mature crisperding technology can not simply be applied in the process of typeset in image processing software; because when directly being applied to existing crisperding technology in the process of typeset; regular meeting changes and makes a mistake; when directly using as Freeman chain code following algorithm following variation may take place, cause the mistake of crisperding in process of typeset:
(1) for the wide target of single pixel, for example straight line though Freeman chain code following algorithm can trace into all pixels, can't turn back to the tracking initiation point, and composing system requires each target area profile all to seal;
(2) because the target area shape that obtains can be very complicated, the two parts that a zone may occur have only the situation of single pixel connection, though can increase some Rule of judgment (as increasing flag bit) this moment in Freeman chain code following algorithm proceeds to follow the tracks of, but the situation less than starting point still might appear following the tracks of back, and increased the complexity of algorithm, reduced efficient;
(3) if the objective contour that obtains does not seal, just the target of having followed the tracks of may be thought when carrying out objective contour tracking next time and still be untreated that such target can obtain two or many profiles, obviously, such user as a result is unacceptable.
Summary of the invention
At the defective that exists in the prior art, the purpose of this invention is to provide a kind of method of in process of typeset, carrying out the image crisperding, this method can easily, efficiently be finished the crisperding of image and handle in process of typeset, make things convenient for the further processing of image, even non-professional image processing personnel also can utilize this method to realize the crisperding processing of image at an easy rate.
For reaching above purpose, the technical solution used in the present invention is: a kind of method of carrying out the image crisperding in process of typeset may further comprise the steps:
(1) image transitions that needs in the process of typeset to handle is become gray level image;
(2) be binary map with the greyscale image transitions in the step (1), the background and the target that are about in the image are carried out image segmentation;
(3) binary map in the step (2) is carried out crisperding and handle, adopt improved Freeman chain code following algorithm to obtain the profile sealing Bezier curve of image object, finish the crisperding of image and handle, concrete grammar is as follows:
1) the picture element matrix equal proportion of image is amplified integral multiple, make single pixel region become many pixel regions;
2) adopt Freeman chain code following algorithm to obtain the profile of target;
3) after all targets in the tracking image, with step 2) in obtain original profile information behind the corresponding multiple of pixel value equal proportion that obtains of Freeman chain code following algorithm.
Further, in the step (1), when image is changed into gray level image,, adopt the gaussian filtering method that image is carried out gaussian filtering according to the tolerance value that the user sets.
Further, when in the step (2) greyscale image transitions being binary map, carry out image transitions according to user's preset threshold.
Further, when in the step (2) greyscale image transitions being binary map, the iteration threshold algorithm can offer approximate threshold value of user, and the general user adopts this threshold value to carry out the result that image transitions can obtain.
Further again, described iteration threshold algorithm may further comprise the steps:
1) intermediate value of getting the gradation of image scope is an initial threshold;
2) be divided into target and background and obtain two integrators according to initial switch function each pixel: target integrator and background integrator input picture;
3) for the first time to after the image scanning end, the value of average two integrators is to determine a new threshold value, utilize this threshold value control switch once more input picture to be divided into target and background, and as new switch function, so iterating no longer change up to switch function; The threshold value that obtain this moment is final threshold value, according to this threshold value gray level image is divided into binary map.
Further, in the step 1) in the step (3), when the picture element matrix equal proportion of image was amplified integral multiple, multiplication factor was 2 times of former picture element matrix.
Further, step 2 in the above-mentioned steps (3)) in, the method of image being carried out inner crisperding is further comprising the steps of: at first be labeled as " not following the tracks of " state with having a few all in the image, to a target following time, each is inserted impact point in the Freeman chain code, it is labeled as " following the tracks of " state, when target following next time, only search is labeled as the point of " following the tracks of " state, thereby avoids repeatedly following the tracks of and can enter into the profile information that target internal obtains inner boundary a profile.
Further, step 2 in the above-mentioned steps (3)) in, the method of image being carried out non-inner crisperding is further comprising the steps of: after adopting Freeman chain code following algorithm to obtain the chain code of a closed area, adopt the Flooding algorithm that all pixels in should the zone in the image all are filled to background colour, seek the starting point of next target area then, carry out profile again and follow the tracks of.
Further, step 2 in the step (3)) in, the method of image being carried out crisperding in the frame is as follows: be the pixel coordinate of image coordinate system with the Coordinate Conversion of have a few on the image on the space of a whole page at first, the pixel coordinate of itself and the rectangular area of image own is asked friendship, the pixel that is positioned in the image outside the common factor is changed to background, keep the pixel value in occuring simultaneously, adopt Freeman chain code following algorithm to carry out the effect that path trace reaches crisperding in the frame to image on this basis at last.
Further, the step 2 in the step (3)) in, the method that image is reversed is as follows: will carry out crisperding again behind the binary map negate look that obtain, thus the effect that obtains reversing.
Effect of the present invention is: adopt method of the present invention, can finish the crisperding of image in process of typeset easily, efficiently handles, and be fit to all kinds of images, make things convenient for the subsequent treatment of image, non-professional image processing personnel also can utilize this method to realize the crisperding processing of image at an easy rate; Simultaneously, thus the user can regulate threshold value according to different demand obtain different crisperding results with tolerance.
Why adopt above-mentioned improved Freeman chain code following algorithm, be because this method has following characteristics: the image pixel matrix equal proportion is amplified integral multiple, make single pixel region become many pixel regions, can handle the wide target of single pixel like this, thereby can directly utilize the Freeman chain code to follow the tracks of; The problem of how having handled when this processing method has avoided existing in the target single pixel wide zone also can guarantee that the profile of each target of obtaining all seals simultaneously; After all targets in having followed the tracks of image,, can not change, influence and lose former profile information with obtaining original profile information behind the corresponding multiple of pixel value scaled down in the Freeman chain code.
Description of drawings
Fig. 1 is the flow chart of the method for the invention embodiment;
Fig. 2 is tolerance value and threshold value selection figure when in process of typeset the image crisperding being handled;
Fig. 3 carries out the design sketch that crisperding is handled in the frame to image in process of typeset;
Fig. 4 carries out the design sketch that crisperding is handled in the non-frame to image in process of typeset;
Fig. 5 carries out the design sketch that inner crisperding is handled to image in process of typeset;
Fig. 6 is to the image design sketch that crisperding handles that reverses in process of typeset;
Fig. 7 carries out the design sketch that crisperding and inner crisperding are handled in the frame simultaneously to image in process of typeset;
Fig. 8 is a design sketch of choosing inner crisperding and counter-rotating to handle simultaneously to image in process of typeset.
Embodiment
Below in conjunction with drawings and the specific embodiments the present invention is described in further detail:
As shown in Figure 1, a kind of method of carrying out the image crisperding in process of typeset comprises the steps:
(1) image transitions that will need to handle in process of typeset becomes gray level image;
When in the present embodiment image being changed into gray level image,, adopt the gaussian filtering method that image is carried out gaussian filtering according to the tolerance value that the user sets.
(2) be binary map with the greyscale image transitions in the step (1), the background and the target that are about in the image are carried out image segmentation;
When greyscale image transitions is binary map, carry out image transitions according to referring to user's preset threshold or the threshold value that adopts the iteration threshold algorithm to obtain, as the threshold value that adopts the iteration threshold algorithm to obtain will make better effects if of the present invention, adopts the iteration threshold algorithm in the present embodiment, and concrete grammar is as follows:
1) intermediate value of getting the gradation of image scope is an initial threshold;
2) be divided into target and background and obtain two integrators according to initial switch function each pixel: target integrator and background integrator input picture;
3) for the first time to after the image scanning end, the value of average two integrators is to determine a new threshold value, utilize this threshold value control switch once more input picture to be divided into target and background, and as new switch function, so iterating no longer change up to switch function; The threshold value that obtain this moment is final threshold value, and gray level image just can be divided into binary map according to this threshold value.
(3) binary map in the step (2) is carried out crisperding and handle, adopt improved Freeman chain code following algorithm to obtain the profile sealing Bezier curve of image object, finish the crisperding of image and handle, concrete grammar is as follows:
1) the picture element matrix equal proportion of image is amplified integral multiple, amplify 2 times in the present embodiment, make single pixel region become many pixel regions;
2) adopt Freeman chain code following algorithm to obtain the profile of target;
3) after all targets in the tracking image, with step 2) in obtain original profile information after one times of the pixel value scaled down that obtains of Freeman chain code following algorithm.
In addition, at the characteristics of software for composing, present embodiment offers the inner crisperding of user, the interior crisperding of frame and three options that reverse simultaneously also can obtain special crisperding effect, and concrete grammar is as follows.
1, inner crisperding:
If do not select inner crisperding, then have to the outermost profile information in target area.If the user has selected inner crisperding, then expression need obtain the information of all target and background intersections in the binary map.For this reason, at first be labeled as " not following the tracks of " state with having a few all in the image.To a target following time, each is inserted impact point in the Freeman chain code, it is labeled as " following the tracks of " state.Like this, when target following next time, only search is labeled as the point of " following the tracks of " state, thereby avoids repeatedly following the tracks of and can enter into the profile information that target internal obtains inner boundary a profile.
2, crisperding in the frame:
Common image is a rectangular area, but in software for composing, and for better artistic effect is arranged, image may enter in the pel of a complexity, and only is presented at the image information of pel inside, can think the image information of having abandoned the pel outside.So if the user has selected crisperding in the frame, then the image crisperding only carries out crisperding to the image in the pel.In order to obtain this effect, be the pixel coordinate of image coordinate system at first with the Coordinate Conversion of have a few on the pel on the space of a whole page, the pixel coordinate of itself and the rectangular area of image own is asked friendship, the pixel that is positioned in the image outside the common factor is changed to background, keep the pixel value in occuring simultaneously.Adopt Freeman chain code following algorithm to carry out the effect that path trace reaches crisperding in the frame to image on this basis.
3, counter-rotating
Because the ambiguity of target, a zone in the image promptly can be considered to target also can be considered to background.At this situation, present embodiment provides the function of counter-rotating, and the binary map inverse that obtains is carried out the effect that crisperding obtains reversing.
Above-mentioned embodiment is applied to actual process of typeset, and it is as follows in process of typeset image to be carried out the effect that crisperding handles:
At first, as shown in Figure 2, enter piece image in the pel of user on the space of a whole page, select or shortcut ejection image crisperding dialog box by menu.Comprise the adjusting of threshold value and tolerance in the dialog box, and the interior crisperding of frame, inner crisperding and counter-rotating OptionButton.Wherein threshold value is the value that adopts the iteration threshold algorithm to obtain according to the image self-information, generally can think the optimal threshold of this image, and the user can directly adopt this threshold value to carry out the image crisperding.The value of tolerance is initially 2.5, according to this value image is carried out gaussian filtering.
The situation of acquiescence is to carry out crisperding in the frame, after the user clicks preview or determines, promptly obtains the crisperding result, as shown in Figure 3.If the user cancels crisperding in the frame, the result who obtains as shown in Figure 4.As can be seen, crisperding is the border with the pel profile in the frame, and the result who obtains is the common factor of pel profile and the profile of image own; But not crisperding then is that entire image is carried out the crisperding operation in the frame.
Fig. 5 has shown the result after the user selects inner crisperding; Fig. 6 has shown counter-rotating back crisperding result.The user can obtain more crisperding effect by the combination of these three functions, and Fig. 7, Fig. 8 are respectively crisperding and the effect of inner crisperding and the results who chooses inner crisperding and counter-rotating to obtain simultaneously in the while check boxes.
Method of the present invention is not limited to the embodiment described in the embodiment, and those skilled in the art's technical scheme according to the present invention draws other execution mode, belongs to technological innovation scope of the present invention equally.
Claims (10)
1. method of carrying out the image crisperding in process of typeset may further comprise the steps:
(1) image transitions that needs in the process of typeset to handle is become gray level image;
(2) be binary map with the greyscale image transitions in the step (1), the background and the target that are about in the image are carried out image segmentation;
(3) binary map in the step (2) is carried out crisperding and handle, adopt improved Freeman chain code following algorithm to obtain the profile sealing Bezier curve of image object, finish the crisperding of image and handle, concrete grammar is as follows:
1) the picture element matrix equal proportion of image is amplified integral multiple, make the wide zone of single pixel become the wide zone of many pixels;
2) adopt Freeman chain code following algorithm to obtain the profile of target;
3) after all targets in the tracking image, with step 2) in obtain original profile information behind the corresponding multiple of profile coordinate scaled down that obtains of Freeman chain code following algorithm.
2. a kind of method of carrying out the image crisperding in process of typeset as claimed in claim 1 is characterized in that: in the step (1), when image is changed into gray level image, according to the tolerance value that the user sets, adopt the gaussian filtering method that image is carried out gaussian filtering.
3. a kind of method of carrying out the image crisperding in process of typeset as claimed in claim 1 is characterized in that: when in the step (2) greyscale image transitions being binary map, carry out image transitions according to user's preset threshold.
4. a kind of method of carrying out the image crisperding in process of typeset as claimed in claim 1 is characterized in that: when in the step (2) greyscale image transitions being binary map, carry out image transitions according to the threshold value that the iteration threshold algorithm obtains.
5. a kind of method of carrying out the image crisperding in process of typeset as claimed in claim 4 is characterized in that: described iteration threshold algorithm may further comprise the steps:
1) intermediate value of getting the gradation of image scope is an initial threshold;
2) be divided into target and background and obtain two integrators according to initial switch function each pixel: target integrator and background integrator input picture;
3) for the first time to after the image scanning end, the value of average two integrators is to determine a new threshold value, utilize this threshold value control switch once more input picture to be divided into target and background, and as new switch function, so iterating no longer change up to switch function; The threshold value that obtain this moment is final threshold value, according to this threshold value gray level image is divided into binary map.
6. as any described a kind of method of carrying out the image crisperding in process of typeset of claim 1 to 5, it is characterized in that: in the step 1) in the step (3), when the picture element matrix equal proportion of image was amplified integral multiple, multiplication factor was 2 times of former picture element matrix.
7. a kind of method of in process of typeset, carrying out the image crisperding as claimed in claim 6, it is characterized in that the step 2 in the step (3)) in, the method of image being carried out inner crisperding is further comprising the steps of: at first be labeled as " not following the tracks of " state with having a few all in the image, to a target following time, each is inserted impact point in the Freeman chain code, it is labeled as " following the tracks of " state, when target following next time, only search is labeled as the point of " following the tracks of " state, thereby avoids repeatedly following the tracks of and can enter into the profile information that target internal obtains inner boundary a profile.
8. a kind of method of in process of typeset, carrying out the image crisperding as claimed in claim 6, it is characterized in that the step 2 in the step (3)) in, the method of image being carried out non-inner crisperding is further comprising the steps of: after adopting Freeman chain code following algorithm to obtain the chain code of a closed area, adopt the Flooding algorithm that all pixels in should the zone in the image all are filled to background colour, seek the starting point of next target area then, carry out profile again and follow the tracks of.
9. a kind of method of in process of typeset, carrying out the image crisperding as claimed in claim 6, it is characterized in that the step 2 in the step (3)) in, the method of image being carried out crisperding in the frame is as follows: be the pixel coordinate of image coordinate system with the Coordinate Conversion of have a few on the image on the space of a whole page at first, the pixel coordinate of itself and the rectangular area of image own is asked friendship, the pixel that is positioned in the image outside the common factor is changed to background, keep the pixel value in occuring simultaneously, adopt Freeman chain code following algorithm to carry out the effect that path trace reaches crisperding in the frame to image on this basis at last.
10. a kind of method of in process of typeset, carrying out the image crisperding as claimed in claim 6, it is characterized in that the step 2 in the step (3)) in, the method that image is reversed is as follows: will carry out crisperding again behind the binary map negate look that obtain, thus the effect that obtains reversing.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB2006100786382A CN100403769C (en) | 2006-04-29 | 2006-04-29 | Image crisperding method during composing process |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB2006100786382A CN100403769C (en) | 2006-04-29 | 2006-04-29 | Image crisperding method during composing process |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN1852389A CN1852389A (en) | 2006-10-25 |
| CN100403769C true CN100403769C (en) | 2008-07-16 |
Family
ID=37133855
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNB2006100786382A Expired - Fee Related CN100403769C (en) | 2006-04-29 | 2006-04-29 | Image crisperding method during composing process |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN100403769C (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101604452B (en) * | 2008-06-13 | 2012-02-01 | 北京大学 | Method for constructing novel closed curve by opened curve and adjacent closed curve |
| CN101916441B (en) * | 2010-08-06 | 2012-02-29 | 西北工业大学 | Freeman chain code-based method for matching curves in digital image |
| CN102542555B (en) * | 2010-12-17 | 2015-06-03 | 北京大学 | Method and system for generating edge seam path and edge seam topological structure of raster image |
| CN109801242B (en) * | 2019-01-23 | 2022-09-02 | 西南交通大学 | Grid image vectorization method based on improved flooding filling algorithm |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN85104598A (en) * | 1985-06-13 | 1986-12-10 | 株式会社写研 | Character processor |
| US5007098A (en) * | 1988-12-30 | 1991-04-09 | Ezel, Inc. | Vectorizing method |
| CN1489114A (en) * | 2003-09-09 | 2004-04-14 | 北京交通大学 | Diveided image coding and decoding method with arbitrary shape region segmentation |
-
2006
- 2006-04-29 CN CNB2006100786382A patent/CN100403769C/en not_active Expired - Fee Related
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN85104598A (en) * | 1985-06-13 | 1986-12-10 | 株式会社写研 | Character processor |
| US5007098A (en) * | 1988-12-30 | 1991-04-09 | Ezel, Inc. | Vectorizing method |
| CN1489114A (en) * | 2003-09-09 | 2004-04-14 | 北京交通大学 | Diveided image coding and decoding method with arbitrary shape region segmentation |
Non-Patent Citations (2)
| Title |
|---|
| 一种基于链码的轮廓平滑算法. 任世宏等.北京理工大学学报，第18卷第4期. 1998 |
| 一种基于链码的轮廓平滑算法. 任世宏等.北京理工大学学报，第18卷第4期. 1998 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN1852389A (en) | 2006-10-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US7609888B2 (en) | Separating a video object from a background of a video sequence | |
| CN102289796B (en) | Interactive image matting | |
| US20150003725A1 (en) | Depth constrained superpixel-based depth map refinement | |
| CN110197153B (en) | Automatic wall identification method in house type graph | |
| EP2431942B1 (en) | Defining a border for an image | |
| Chen et al. | A novel color edge detection algorithm in RGB color space | |
| CN102782706A (en) | Text enhancement of a textual image undergoing optical character recognition | |
| CN102999901A (en) | Method and system for processing split online video on the basis of depth sensor | |
| KR102010085B1 (en) | Method and apparatus for producing labeling image of microstructure using super-pixels | |
| CN100403769C (en) | Image crisperding method during composing process | |
| CN110675463B (en) | Method and device for generating partial color pencil drawing | |
| CN115147448A (en) | Image enhancement and feature extraction method for automatic welding | |
| CN102324099A (en) | Step edge detection method oriented to humanoid robot | |
| CN104021566A (en) | GrabCut algorithm-based automatic segmentation method of tongue diagnosis image | |
| US8929587B2 (en) | Multi-tracking of video objects | |
| CN103942756B (en) | A kind of method of depth map post processing and filtering | |
| KR101028699B1 (en) | Apparatus and method for painterly rendering | |
| CN105373798A (en) | K neighbor image matting and mathematical morphology-based calligraphy character extracting method | |
| JP3716455B2 (en) | Region extraction method and region extraction device | |
| CN107610126A (en) | A kind of interactive image segmentation method based on local prior distribution | |
| CN108090895B (en) | Container lockhole contour extraction method based on image processing | |
| CN101799931B (en) | Painting rendering method based on colour feature study | |
| Masood et al. | Capturing outlines of 2D objects with Bézier cubic approximation | |
| US8942491B2 (en) | Topology-preserving downsampling of binary images | |
| Yagi | A filter based approach for inbetweening |
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 | ||
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20080716 Termination date: 20190429 |