CN104794472A - Method and device used for extracting gesture edge image and gesture extracting method - Google Patents
Method and device used for extracting gesture edge image and gesture extracting method Download PDFInfo
- Publication number
- CN104794472A CN104794472A CN201410024791.1A CN201410024791A CN104794472A CN 104794472 A CN104794472 A CN 104794472A CN 201410024791 A CN201410024791 A CN 201410024791A CN 104794472 A CN104794472 A CN 104794472A
- Authority
- CN
- China
- Prior art keywords
- gesture
- image
- edge image
- edge
- pixel
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Abstract
The invention relates to a method and a device used for extracting a gesture edge image and a gesture extracting method. The method used for extracting the gesture edge image comprises the steps of acquiring a binarized foreground edge image based on an image taking a gesture as the foreground; scaling a predetermined gesture edge template image according to the size of the gesture in the image taking the gesture as the foreground; dividing the scaled gesture template image into a plurality of sub-portions, and respectively selecting corresponding image blocks in the foreground edge image by using each sub-portion; and acquiring a portion of the gesture edge image from each image block, thereby extracting the whole gesture edge image.
Description
Technical field
The present invention relates to a kind of method and apparatus for extracting gesture edge image and a kind of gesture extracting method.
Background technology
In man-machine interaction, gesture is a kind of important medium, and such as, gesture estimation of Depth in three dimensions can provide technical support at the application scenarios of various opening.And only having in the scene of monocular-camera, the Accurate Segmentation of gesture plays a key effect to the estimation of the gesture degree of depth.
Traditional Hand Gesture Segmentation method comprises the color utilizing gauss hybrid models (GMM) to come matching gesture and background, and then is distinguished the two by color classification.Meanwhile, method is also had to utilize the color of color histogram to gesture and background to distinguish.In addition, some methods based on active contour model utilize sample training to obtain general gesture shape, and initial gesture shape basis is obtained by Local Search the exact boundary of palm.In addition, the heterochromia of the figure segmentation method color model and image zones of different that have merged gauss hybrid models matching is split target.
But in complex environment, if background and the colour of skin are very close, then said method all will lose efficacy.
For the defect existed in prior art, the application is proposed.
Summary of the invention
Provide hereinafter about brief overview of the present invention, to provide about the basic comprehension in some of the present invention.Should be appreciated that this general introduction is not summarize about exhaustive of the present invention.It is not that intention determines key of the present invention or pith, and nor is it intended to limit the scope of the present invention.Its object is only provide some concept in simplified form, in this, as the preorder in greater detail discussed after a while.
A fundamental purpose of the present invention is to provide a kind of method and apparatus for extracting gesture edge image and a kind of gesture extracting method, at least to overcome existing problem.
According to an aspect of the present invention, a kind of method for extracting gesture edge image is provided.The method comprises: based on the image taking gesture as prospect, obtains the foreground edge image of binaryzation; According to the size of the gesture in the described image taking gesture as prospect, predetermined gesture border template image is carried out convergent-divergent; Gesture border template image after convergent-divergent is divided into multiple subdivision, and uses each subdivision to select the image block of the correspondence in described foreground edge image respectively; And from each image block, obtain a part for gesture edge image respectively, thus extract whole gesture edge image.
According to another aspect of the present invention, a kind of gesture extracting method is provided.This comprises: by the method for extracting gesture edge image described in the embodiment of the present invention, from gesture be prospect image extract gesture edge image; And utilize figure to cut theoretical frame, extract gesture in described image by arranging different weights to gesture edge image and other edge image.
According to a further aspect of the invention, a kind of device for extracting gesture edge image is provided.This device comprises: acquiring unit, and it is configured to the image based on taking gesture as prospect, obtains the foreground edge image of binaryzation; Unit for scaling, it is configured to, according to the size of the gesture in the described image taking gesture as prospect, predetermined gesture border template image be carried out convergent-divergent; Division unit, it is configured to the gesture border template image after by convergent-divergent and is divided into multiple subdivision, and uses each subdivision to select the image block of the correspondence in described foreground edge image respectively; And extraction unit, it is configured to extraction unit, and it is configured to the part obtaining gesture edge image from each image block respectively, thus extracts whole gesture edge image.
In addition, embodiments of the invention additionally provide the computer program for realizing said method.
In addition, embodiments of the invention additionally provide the computer program of at least computer-readable medium form, it record the computer program code for realizing said method.
According to embodiments of the invention, utilize shape prior to instruct the segmentation carrying out gesture and background.Even if so in background and the very close complex environment of the colour of skin, also can effectively distinguish colour of skin edge and background edge, thus effectively can carry out the segmentation of gesture and background.
By below in conjunction with the detailed description of accompanying drawing to most preferred embodiment of the present invention, these and other advantage of the present invention will be more obvious.
Accompanying drawing explanation
Below with reference to the accompanying drawings illustrate embodiments of the invention, above and other objects, features and advantages of the present invention can be understood more easily.Parts in accompanying drawing are just in order to illustrate principle of the present invention.In the accompanying drawings, same or similar technical characteristic or parts will adopt same or similar Reference numeral to represent.
Fig. 1 is the process flow diagram of the method for extracting gesture edge image schematically shown according to the embodiment of the present invention;
Fig. 2 illustrates with gesture to be the image of prospect and the schematic diagram of the foreground image of the binaryzation obtained;
Fig. 3 is the schematic diagram that foreground edge image is shown;
Fig. 4 is the schematic diagram of the subdivision after gesture border template image, its subdivision and range conversion are shown;
Fig. 5 be illustrate whether carry out sign reversing and produce pixel value distribution schematic diagram;
Fig. 6 illustrates by processing the schematic diagram obtaining whole gesture edge image to each subdivision;
Fig. 7 schematically shows according to an embodiment of the invention for extracting the block diagram of the device of gesture edge image;
Fig. 8 is the block diagram schematically showing extraction unit according to an embodiment of the invention;
Fig. 9 is the block diagram schematically showing division unit according to an embodiment of the invention;
Figure 10 is the block diagram schematically showing acquiring unit according to an embodiment of the invention;
Figure 11 shows the structural drawing of the citing that may be used for the computing equipment implementing method and apparatus for extracting gesture edge image of the present invention and gesture extracting method.
Embodiment
With reference to the accompanying drawings embodiments of the invention are described.The element described in an accompanying drawing of the present invention or a kind of embodiment and feature can combine with the element shown in one or more other accompanying drawing or embodiment and feature.It should be noted that for purposes of clarity, accompanying drawing and eliminate expression and the description of unrelated to the invention, parts known to persons of ordinary skill in the art and process in illustrating.
Fig. 1 is the process flow diagram of the method for extracting gesture edge image schematically shown according to the embodiment of the present invention.Referring to Fig. 1 describe according to the embodiment of the present invention for extracting gesture edge image.
In step S101, based on the image taking gesture as prospect, obtain the foreground edge image of binaryzation.
First, the image of the hand region in arbitrary image can be obtained by the cascade classifier that training in advance is good, as the image taking gesture as prospect.Such as, obtain with gesture be the image of prospect as shown in Figure 2 (a) shows.
Based on taking gesture as the image of prospect such as shown in Fig. 2 (a), carrying out prospect estimation by Vibe algorithm, thus obtaining the foreground image of binaryzation.Fig. 2 (b) shows the foreground image V of obtained binaryzation, and white point is wherein the foreground pixel point estimated.From as shown in Figure 2 (a) shows be extract edge image E the image of prospect with gesture.Edge image E and the foreground image V of the binaryzation obtained are carried out and operation, thus obtain foreground edge image Fe:
Fe=E AND V。
In order to the noise of the non-flesh tone portion of filtering, the skin color model that also use training in advance is good carrys out the noise edge in filtering foreground edge image as mask, foreground edge image Fe and skin color mask S is carried out and operation, obtains the foreground edge F:F=Fe AND S of area of skin color.The filtered foreground edge image F obtained as shown in Figure 3.
In step s 102, according to gesture be prospect image in the size of gesture, predetermined gesture border template image is carried out convergent-divergent, with make the gesture border template image after convergent-divergent with gesture be prospect image in the size of gesture match.As shown in Figure 4 (a), wherein, with gesture be prospect image in the size of gesture matches is the brighter white gesture border template image of color.
In step s 103, the gesture border template image after convergent-divergent is divided into multiple subdivision, and uses each subdivision to select the image block of the correspondence in described foreground edge image F respectively.
Particularly, the edge pixel point (white pixel point) in the gesture border template image after convergent-divergent is encoded to continuous coordinate array.Such as, can by extracting the coordinate of starting point in the enterprising line scanning of gesture border template image.And can Depth Priority Searching be used, obtain continuous coordinate array, such as, by searching for 8 neighborhood points using starting point as input point, and by 8 neighborhood points, also the point be not included in continuous array be included in continuous array, the neighborhood point be included in continuous array is then used as input point to carry out recursion backtrack search, thus obtains described continuous coordinate array.
Then, obtained continuous coordinate array is divided into multiple coordinate subnumber group.Such as, by scanning in x direction, and can scan in y direction, searching for the minimum and maximum coordinate of each subnumber group, thus obtain the boundary rectangle enclosing subdivision array respectively, with this, continuous array is divided into multiple subdivision.
Above-mentioned method gesture border template image after convergent-divergent being divided into multiple subdivision is only exemplary, and those skilled in the art it will be appreciated that various ways is to carry out the division of above-mentioned gesture border template image in addition.
In step S104, from each image block, obtain a part for gesture edge image respectively, thus extract whole gesture edge image.
The image after conversion is obtained by carrying out range conversion to gesture border template image.Range conversion is a kind of mode effectively comparing two point set distances.In order to obtain the difference of foreground edge image F and gesture template edge image, we carry out range conversion to gesture border template image.Then, the value of the pixel of corresponding foreground edge image F can read from the image after range conversion, and it represents the distance value of foreground edge and gesture template edge.Fig. 4 (b) shows in foreground edge image, corresponding with a subdivision (in Fig. 4 (a) bottom-right part) of gesture border template image image block.Fig. 4 (c) shows the gesture border template image after range conversion, and the grey lines from top to bottom being sideling positioned at Fig. 4 (c) center are the gesture border template image after range conversion.
In order to ensure the continuity of the edge pixel values being positioned at gesture border template fringe region, as shown in Fig. 4 (d), if the closed gesture graph that the pixel in foreground edge image is in the gesture border template image after convergent-divergent is inner, then the pixel value in this pixel gesture border template image be after the conversion on the occasion of; If the closed gesture graph that the pixel in foreground edge image is in the gesture border template image after convergent-divergent is outside, then the pixel value in this pixel gesture border template image is after the conversion negative value.Left diagram in Fig. 5 and right diagram represent respectively whether carry out this kind of sign reversing and produce pixel value distribution.In left diagram in Figure 5, owing to having carried out this kind of sign reversing, the edge pixel values being therefore positioned at gesture border template fringe region has been continuous print.And in right diagram in Figure 5, owing to not carrying out this kind of sign reversing, the edge pixel values being therefore positioned at gesture border template fringe region is discontinuous.
Pixel value in statistics subdivision, according to the threshold value preset, the pixel larger to the pixel value in the image after conversion carries out filtering, because these pixels and gesture border template distance are comparatively greatly, is not therefore the pixel of formation prospect gesture.The pixel that pixel value is larger is such as those pixels that pixel value is greater than the threshold value preset.To carry out cluster except processing rear remaining pixel (white dashed line as shown in Figure 6 (a) represents) after filtration, gesture edge image has similar distance in the gesture border template after range conversion, and therefore they can form maximum cluster.As shown in Figure 6 (b), obtained a part for gesture edge image by the multiple pixels forming maximum cluster, this part corresponds to the gesture edge image portion in Fig. 6 (c) lower right box.A part for rest of pixels point then edge image as a setting.
Then similarly, then to the next subdivision (subdivision that the upper right box such as, in Fig. 6 (d) represents) in foreground edge image above-mentioned process is carried out.After all above-mentioned process is carried out to all subdivisions, whole gesture edge image (as shown in Figure 6 (e)) can be obtained.
According to the method for extracting gesture edge image of the embodiment of the present invention, utilize shape prior to instruct the extraction carrying out gesture edge image.Even if so in background and the very close complex environment of the colour of skin, also can effectively extract gesture edge image.
According to a further aspect in the invention, a kind of gesture extracting method is provided.An embodiment of gesture extracting method is by the above-described method for extracting gesture edge image, from gesture be prospect image extract gesture edge image.Then, utilize figure segmentation framework, extract gesture in image by arranging different weights according to above-mentioned gesture edge image edge part and non-edge part.Namely think that prospect gesture edge is strong ties, and background edge more should by the border as segmentation.Therefore can arrange prospect gesture edge conjunction weights is 1, and background edge connection weights are 0.Meanwhile, they also can be set to the connection weights with the pixel of its neighbour.Then, optimum segmentation border is searched for by maximum-flow algorithm.
In order to filtering figure cuts the possible non-colour of skin noise after process, the good skin color model of training in advance can also be used to carry out the noise in the gesture that filtering extracts as mask.Thus the palm image of expectation can be obtained from image.
According to the gesture extracting method of the embodiment of the present invention, utilize shape prior to instruct the extraction carrying out gesture.Even if so in background and the very close complex environment of the colour of skin, also can effectively distinguish colour of skin edge and background edge, thus effectively can carry out gesture extraction.
Describe according to an embodiment of the invention for extracting the device 700 of gesture edge image referring to Fig. 7.
Fig. 7 schematically shows according to an embodiment of the invention for extracting the block diagram of the device 700 of gesture edge image.Wherein, illustrate only with the present invention closely-related part for brevity.For extracting in the device 700 of gesture edge image, the method for extracting gesture edge image described by above reference diagram 1 can be performed.As shown in Figure 7, the device 700 for extracting gesture edge image can comprise acquiring unit 701, unit for scaling 702, division unit 703 and extraction unit 704.
Acquiring unit 701 can be configured to the image based on taking gesture as prospect, obtains the foreground edge image of binaryzation.Unit for scaling 702 can be configured to, according to the size of the gesture in the described image taking gesture as prospect, predetermined gesture border template image be carried out convergent-divergent.Division unit 703 can be configured to the gesture border template image after by convergent-divergent and be divided into multiple subdivision, and uses each subdivision to select the image block of the correspondence in described foreground edge image respectively.Extraction unit 704 can be configured to the part obtaining gesture edge image from each image block respectively, thus extracts whole gesture edge image.
Particularly, the edge pixel point that division unit 703 can be configured in the gesture border template image after by convergent-divergent is further encoded to continuous coordinate array, described continuous coordinate array is divided into multiple coordinate subnumber group, and obtains each subdivision by the minimum and maximum coordinate searching for each subnumber group.
According to an embodiment of the invention for extracting the device of gesture edge image, passable
Referring to Fig. 8, the device for extracting gesture edge image is according to another embodiment of the invention described.
Fig. 8 is the block diagram schematically showing extraction unit according to an embodiment of the invention.As shown in Figure 8, the device for extracting gesture edge image according to an embodiment of the invention is except comprising the unit shown in Fig. 7, and extraction unit 704 may further include: range conversion portion 801, pixel filtering portion 802 and cluster portion 803.
Range conversion portion 801 can be configured to carry out range conversion to gesture border template image.Pixel filtering portion 802 can be configured to according to the threshold value that presets, by the image after conversion, pixel that pixel value in corresponding with image block subdivision is larger carries out filtering.Cluster portion 803 can be configured to, by after filtration except the remaining pixel in the subdivision of the described correspondence after process carries out cluster, be obtained a part for gesture edge image by the multiple pixels forming maximum cluster.
Referring to Fig. 9, the device for extracting gesture edge image is according to another embodiment of the invention described.
Fig. 9 is the block diagram schematically showing division unit according to an embodiment of the invention.As shown in Figure 9, according to another embodiment of the invention for extracting the device of gesture edge image except comprising the unit shown in Fig. 7, division unit 703 may further include: starting point coordinate extraction unit 901 and coordinate array acquisition unit 902.
Starting point coordinate extraction unit 901 can be configured to the coordinate by extracting starting point in the enterprising line scanning of gesture border template image.Coordinate array acquisition unit 902 can be configured to use Depth Priority Searching, by searching for 8 neighborhood points using starting point as input point, and by 8 neighborhood points, also the point be not included in continuous array be included in continuous array, the neighborhood point be included in continuous array is then used as input point to carry out recursion backtrack search, thus obtains described continuous coordinate array.
Referring to Figure 10, the device for extracting gesture edge image is according to another embodiment of the invention described.
Figure 10 is the block diagram schematically showing acquiring unit according to an embodiment of the invention.As shown in Figure 10, according to another embodiment of the invention for extracting the device of gesture edge image except comprising the unit shown in Fig. 7, acquiring unit 701 may further include: foreground image acquisition unit 1001, edge image extraction unit 1002 and foreground edge image acquiring unit 1003.
Foreground image acquisition unit 1001 can be configured to the image based on taking gesture as prospect, is obtained the foreground image of binaryzation by Vibe algorithm.Edge image extraction unit 1002 can be configured to from described be extract edge image the image of prospect with gesture.Foreground edge image acquiring unit 1003 can be configured to extracted described edge image and described foreground image be carried out with computing to obtain described foreground edge image.
According to the device for extracting gesture edge image of the embodiment of the present invention, utilize shape prior to instruct the extraction carrying out gesture edge image.Even if so in background and the very close complex environment of the colour of skin, also can effectively extract gesture edge image.
Below ultimate principle of the present invention is described in conjunction with specific embodiments, but, it is to be noted, for those of ordinary skill in the art, whole or any step or the parts of method and apparatus of the present invention can be understood, can in the network of any calculation element (comprising processor, storage medium etc.) or calculation element, realized with hardware, firmware, software or their combination, this is that those of ordinary skill in the art use their basic programming skill just can realize when having read explanation of the present invention.
Therefore, object of the present invention can also be realized by an operation program or batch processing on any calculation element.Calculation element can be known fexible unit.Therefore, object of the present invention also can realize only by the program product of providing package containing the program code of implementation method or device.That is, such program product also forms the present invention, and the storage medium storing such program product also forms the present invention.Obviously, storage medium can be any storage medium developed in any known storage medium or future.
When realizing embodiments of the invention by software and/or firmware, from storage medium or network to the computing machine with specialized hardware structure, the program forming this software installed by multi-purpose computer 1100 such as shown in Figure 11, this computing machine, when being provided with various program, can perform various function etc.
In fig. 11, CPU (central processing unit) (CPU) 1101 performs various process according to the program stored in ROM (read-only memory) (ROM) 1102 or from the program that storage area 1108 is loaded into random access memory (RAM) 1103.In RAM1103, also store the data required when CPU1101 performs various process etc. as required.CPU1101, ROM1102 and RAM1103 are via bus 1104 link each other.Input/output interface 1105 also link to bus 1104.
Following parts link is to input/output interface 1105: importation 1106(comprises keyboard, mouse etc.), output 1107(comprises display, such as cathode-ray tube (CRT) (CRT), liquid crystal display (LCD) etc., and loudspeaker etc.), storage area 1108(comprises hard disk etc.), communications portion 1109(comprises network interface unit such as LAN card, modulator-demodular unit etc.).Communications portion 1109 is via network such as the Internet executive communication process.As required, driver 1110 also can link to input/output interface 1105.Detachable media 1111 such as disk, CD, magneto-optic disk, semiconductor memory etc. are installed on driver 1110 as required, and the computer program therefrom read is installed in storage area 1108 as required.
When series of processes above-mentioned by software simulating, from network such as the Internet or storage medium, such as detachable media 1111 installs the program forming software.
It will be understood by those of skill in the art that this storage medium is not limited to wherein having program stored therein shown in Figure 11, distributes the detachable media 1111 to provide program to user separately with equipment.The example of detachable media 1111 comprises disk (comprising floppy disk (registered trademark)), CD (comprising compact disc read-only memory (CD-ROM) and digital universal disc (DVD)), magneto-optic disk (comprising mini-disk (MD) (registered trademark)) and semiconductor memory.Or hard disk that storage medium can be ROM1102, comprise in storage area 1108 etc., wherein computer program stored, and user is distributed to together with comprising their equipment.
The present invention also proposes a kind of program product storing the instruction code of machine-readable.When instruction code is read by machine and performs, the above-mentioned method according to the embodiment of the present invention can be performed.
Correspondingly, be also included within of the present invention disclosing for carrying the above-mentioned storage medium storing the program product of the instruction code of machine-readable.Storage medium includes but not limited to floppy disk, CD, magneto-optic disk, storage card, memory stick etc.
In addition, some is for the method and apparatus of combination tool according to an embodiment of the invention, can expand the usable range of combination.
Those of ordinary skill in the art should be understood that what exemplify at this is exemplary, and the present invention is not limited thereto.
As an example, each step of said method and all modules of the said equipment and/or unit may be embodied as software, firmware, hardware or its combination, and as the part in relevant device.When in said apparatus, all modules, unit are configured by software, firmware, hardware or its mode combined, spendable concrete means or mode are well known to those skilled in the art, and do not repeat them here.
As an example, when being realized by software or firmware, to the computing machine (multi-purpose computer 1100 such as shown in Figure 11) with specialized hardware structure, the program forming this software can be installed from storage medium or network, this computing machine, when being provided with various program, can perform various functions etc.
Above in the description of the specific embodiment of the invention, the feature described for a kind of embodiment and/or illustrate can use in one or more other embodiment in same or similar mode, combined with the feature in other embodiment, or substitute the feature in other embodiment.
Should emphasize, term " comprises/comprises " existence referring to feature, key element, step or assembly when using herein, but does not get rid of the existence or additional of one or more further feature, key element, step or assembly.
In addition, method of the present invention be not limited to specifications in describe time sequencing perform, also can according to other time sequencing ground, perform concurrently or independently.Therefore, the execution sequence of the method described in this instructions is not construed as limiting technical scope of the present invention.
Although above by the description of specific embodiments of the invention to invention has been disclosure, but, should be appreciated that, those skilled in the art can design various amendment of the present invention, improvement or equivalent in the spirit and scope of claims.These amendments, improvement or equivalent also should be believed to comprise in protection scope of the present invention.
About the embodiment comprising above embodiment, following remarks is also disclosed:
Remarks 1. 1 kinds, for extracting the method for gesture edge image, comprising:
Based on the image taking gesture as prospect, obtain the foreground edge image of binaryzation;
According to the size of the gesture in the described image taking gesture as prospect, predetermined gesture border template image is carried out convergent-divergent;
Gesture border template image after convergent-divergent is divided into multiple subdivision, and uses each subdivision to select the image block of the correspondence in described foreground edge image respectively; And
From each image block, obtain a part for gesture edge image respectively, thus extract whole gesture edge image.
The method of remarks 2. according to remarks 1, wherein, describedly gesture border template image after convergent-divergent is divided into multiple subdivision comprises step: the edge pixel point in the gesture border template image after convergent-divergent is encoded to continuous coordinate array, described continuous coordinate array is divided into multiple coordinate subnumber group, and obtains each subdivision by the minimum and maximum coordinate searching for each subnumber group.
The method of remarks 3. according to remarks 1 or 2, wherein, the described part obtaining gesture edge image from each image block respectively comprises the following steps:
Range conversion is carried out to gesture border template image;
According to the threshold value preset, by the image after conversion, pixel that pixel value in corresponding with image block subdivision is larger carries out filtering; And
By after filtration except the remaining pixel in the subdivision of the described correspondence after process carries out cluster, obtained a part for gesture edge image by the multiple pixels forming maximum cluster.
The method of remarks 4. according to remarks 2, wherein, is describedly encoded to continuous coordinate array by the edge pixel point in the gesture border template image after convergent-divergent and comprises the following steps:
By extracting the coordinate of starting point in the enterprising line scanning of gesture border template image; And
Use Depth Priority Searching, by searching for 8 neighborhood points using starting point as input point, and by 8 neighborhood points, also the point be not included in continuous array be included in continuous array, the neighborhood point be included in continuous array is then used as input point to carry out recursion backtrack search, thus obtains described continuous coordinate array.
The method of remarks 5. according to remarks 1 or 2, wherein, the described image based on taking gesture as prospect, the foreground edge image obtaining binaryzation comprises the following steps:
Based on the image taking gesture as prospect, obtained the foreground image of binaryzation by Vibe algorithm;
From described be extract edge image the image of prospect with gesture; And
Extracted described edge image and described foreground image are carried out with computing to obtain described foreground edge image.
The method of remarks 6. according to remarks 5, wherein, obtains the hand region image in coloured image by the cascade classifier that training in advance is good, as described take gesture as the image of prospect.
The method of remarks 7. according to remarks 5, wherein, the skin color model that also use training in advance is good carrys out the noise edge in foreground edge image described in filtering as mask.
The method of remarks 8. according to remarks 3, wherein,
If the closed gesture graph that the pixel in described foreground edge image is in the gesture border template image after convergent-divergent is inner, then the pixel value in the image of this pixel after described conversion be on the occasion of;
If the closed gesture graph that the pixel in described foreground edge image is in the gesture border template image after convergent-divergent is outside, then the pixel value in the image of this pixel after described conversion is negative value.
Remarks 9. 1 kinds of gesture extracting method, comprising:
By the method for extracting gesture edge image according to above-mentioned remarks, from gesture be prospect image extract gesture edge image; And
Utilize figure to cut theoretical frame, extract gesture in described image by arranging different weights to gesture edge image and other edge image.
The method of remarks 11. according to remarks 9, wherein, also uses the good skin color model of training in advance to carry out the noise in the gesture that filtering extracts as mask.
Remarks 12. 1 kinds, for extracting the device of gesture edge image, comprising:
Acquiring unit, it is configured to the image based on taking gesture as prospect, obtains the foreground edge image of binaryzation;
Unit for scaling, it is configured to, according to the size of the gesture in the described image taking gesture as prospect, predetermined gesture border template image be carried out convergent-divergent;
Division unit, it is configured to the gesture border template image after by convergent-divergent and is divided into multiple subdivision, and uses each subdivision to select the image block of the correspondence in described foreground edge image respectively; And
Extraction unit, it is configured to the part obtaining gesture edge image from each image block respectively, thus extracts whole gesture edge image.
The device of remarks 13. according to remarks 12, wherein, the edge pixel point that described division unit is configured in the gesture border template image after by convergent-divergent is further encoded to continuous coordinate array, described continuous coordinate array is divided into multiple coordinate subnumber group, and obtains each subdivision by the minimum and maximum coordinate searching for each subnumber group.
The device of remarks 14. according to remarks 12 or 13, wherein, described extraction unit comprises further:
Range conversion portion, it is configured to carry out range conversion to gesture border template image;
Pixel filtering portion, it is configured to according to the threshold value that presets, by the image after conversion, pixel that pixel value in corresponding with image block subdivision is larger carries out filtering; And
Cluster portion, it is configured to, by after filtration except the remaining pixel in the subdivision of the described correspondence after process carries out cluster, be obtained a part for gesture edge image by the multiple pixels forming maximum cluster.
The device of remarks 15. according to remarks 13, wherein, described division unit comprises further:
Starting point coordinate extraction unit, it is configured to the coordinate by extracting starting point in the enterprising line scanning of gesture border template image; And
Coordinate array acquisition unit, it is configured to use Depth Priority Searching, by searching for 8 neighborhood points using starting point as input point, and by 8 neighborhood points, also the point be not included in continuous array be included in continuous array, the neighborhood point be included in continuous array is then used as input point to carry out recursion backtrack search, thus obtains described continuous coordinate array.
The device of remarks 16. according to remarks 12 or 13, wherein, described acquiring unit comprises further:
Foreground image acquisition unit, it is configured to the image based on taking gesture as prospect, is obtained the foreground image of binaryzation by Vibe algorithm;
Edge image extraction unit, its be configured to from described be extract edge image the image of prospect with gesture; And
Foreground edge image acquiring unit, it is configured to extracted described edge image and described foreground image be carried out with computing to obtain described foreground edge image.
Claims (10)
1., for extracting a method for gesture edge image, comprising:
Based on the image taking gesture as prospect, obtain the foreground edge image of binaryzation;
According to the size of the gesture in the described image taking gesture as prospect, predetermined gesture border template image is carried out convergent-divergent;
Gesture border template image after convergent-divergent is divided into multiple subdivision, and uses each subdivision to select the image block of the correspondence in described foreground edge image respectively; And
From each image block, obtain a part for gesture edge image respectively, thus extract whole gesture edge image.
2. method according to claim 1, wherein, describedly gesture border template image after convergent-divergent is divided into multiple subdivision comprises step: the edge pixel point in the gesture border template image after convergent-divergent is encoded to continuous coordinate array, described continuous coordinate array is divided into multiple coordinate subnumber group, and obtains each subdivision by the minimum and maximum coordinate searching for each subnumber group.
3. method according to claim 1 and 2, wherein, the described part obtaining gesture edge image from each image block respectively comprises the following steps:
Range conversion is carried out to gesture border template image;
According to the threshold value preset, by the image after conversion, pixel that pixel value in corresponding with image block subdivision is larger carries out filtering; And
By after filtration except the remaining pixel in the subdivision of the described correspondence after process carries out cluster, obtained a part for gesture edge image by the multiple pixels forming maximum cluster.
4. method according to claim 2, wherein, is describedly encoded to continuous coordinate array by the edge pixel point in the gesture border template image after convergent-divergent and comprises the following steps:
By extracting the coordinate of starting point in the enterprising line scanning of gesture border template image; And
Use Depth Priority Searching, by searching for 8 neighborhood points using starting point as input point, and by 8 neighborhood points, also the point be not included in continuous array be included in continuous array, the neighborhood point be included in continuous array is then used as input point to carry out recursion backtrack search, thus obtains described continuous coordinate array.
5. method according to claim 1 and 2, wherein, the described image based on taking gesture as prospect, the foreground edge image obtaining binaryzation comprises the following steps:
Based on the image taking gesture as prospect, obtained the foreground image of binaryzation by Vibe algorithm;
From described be extract edge image the image of prospect with gesture; And
Extracted described edge image and described foreground image are carried out with computing to obtain described foreground edge image.
6. method according to claim 3, wherein,
If the closed gesture graph that the pixel in described foreground edge image is in the gesture border template image after convergent-divergent is inner, then the pixel value in the image of this pixel after described conversion be on the occasion of;
If the closed gesture graph that the pixel in described foreground edge image is in the gesture border template image after convergent-divergent is outside, then the pixel value in the image of this pixel after described conversion is negative value.
7. a gesture extracting method, comprising:
By the method for extracting gesture edge image according to the claims, from gesture be prospect image extract gesture edge image; And
Utilize figure to cut theoretical frame, extract gesture in described image by arranging different weights to gesture edge image and other edge image.
8., for extracting a device for gesture edge image, comprising:
Acquiring unit, it is configured to the image based on taking gesture as prospect, obtains the foreground edge image of binaryzation;
Unit for scaling, it is configured to, according to the size of the gesture in the described image taking gesture as prospect, predetermined gesture border template image be carried out convergent-divergent;
Division unit, it is configured to the gesture border template image after by convergent-divergent and is divided into multiple subdivision, and uses each subdivision to select the image block of the correspondence in described foreground edge image respectively; And
Extraction unit, it is configured to the part obtaining gesture edge image from each image block respectively, thus extracts whole gesture edge image.
9. device according to claim 8, wherein, the edge pixel point that described division unit is configured in the gesture border template image after by convergent-divergent is further encoded to continuous coordinate array, described continuous coordinate array is divided into multiple coordinate subnumber group, and obtains each subdivision by the minimum and maximum coordinate searching for each subnumber group.
10. device according to claim 8 or claim 9, wherein, described extraction unit comprises further:
Range conversion portion, it is configured to carry out range conversion to gesture border template image;
Pixel filtering portion, it is configured to according to the threshold value that presets, by the image after conversion, pixel that pixel value in corresponding with image block subdivision is larger carries out filtering; And
Cluster portion, it is configured to, by after filtration except the remaining pixel in the subdivision of the described correspondence after process carries out cluster, be obtained a part for gesture edge image by the multiple pixels forming maximum cluster.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410024791.1A CN104794472A (en) | 2014-01-20 | 2014-01-20 | Method and device used for extracting gesture edge image and gesture extracting method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410024791.1A CN104794472A (en) | 2014-01-20 | 2014-01-20 | Method and device used for extracting gesture edge image and gesture extracting method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN104794472A true CN104794472A (en) | 2015-07-22 |
Family
ID=53559259
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410024791.1A Pending CN104794472A (en) | 2014-01-20 | 2014-01-20 | Method and device used for extracting gesture edge image and gesture extracting method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104794472A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105957059A (en) * | 2016-04-20 | 2016-09-21 | 广州视源电子科技股份有限公司 | Electronic component missing detection method and system |
| CN106682627A (en) * | 2016-12-29 | 2017-05-17 | 北京奇虎科技有限公司 | Identifying method and device of palm print data |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5168529A (en) * | 1988-08-29 | 1992-12-01 | Rayethon Company | Confirmed boundary pattern matching |
| CN102368290A (en) * | 2011-09-02 | 2012-03-07 | 华南理工大学 | Hand gesture identification method based on finger advanced characteristic |
-
2014
- 2014-01-20 CN CN201410024791.1A patent/CN104794472A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5168529A (en) * | 1988-08-29 | 1992-12-01 | Rayethon Company | Confirmed boundary pattern matching |
| CN102368290A (en) * | 2011-09-02 | 2012-03-07 | 华南理工大学 | Hand gesture identification method based on finger advanced characteristic |
Non-Patent Citations (1)
| Title |
|---|
| 顾理 等: "一种基于模板匹配的手形认证算法", 《计算机工程与应用》 * |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105957059A (en) * | 2016-04-20 | 2016-09-21 | 广州视源电子科技股份有限公司 | Electronic component missing detection method and system |
| CN105957059B (en) * | 2016-04-20 | 2019-03-01 | 广州视源电子科技股份有限公司 | Electronic component missing part detection method and system |
| CN106682627A (en) * | 2016-12-29 | 2017-05-17 | 北京奇虎科技有限公司 | Identifying method and device of palm print data |
| CN106682627B (en) * | 2016-12-29 | 2021-01-29 | 北京奇虎科技有限公司 | Palm print data identification method and device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110738207B (en) | Character detection method for fusing character area edge information in character image | |
| Dalla Mura et al. | Classification of hyperspectral images by using extended morphological attribute profiles and independent component analysis | |
| CN111145209B (en) | Medical image segmentation method, device, equipment and storage medium | |
| CN108416268B (en) | Action recognition method based on double-robot visual communication | |
| JP2006246435A5 (en) | ||
| CN107784301A (en) | Method and apparatus for identifying character area in image | |
| JPH03218581A (en) | Picture segmentation method | |
| US20210248729A1 (en) | Superpixel merging | |
| CN108830780A (en) | Image processing method and device, electronic equipment, storage medium | |
| CN105719243A (en) | Image processing device and method | |
| Fu et al. | Robust image segmentation using contour-guided color palettes | |
| CN104021566A (en) | GrabCut algorithm-based automatic segmentation method of tongue diagnosis image | |
| CN105139342A (en) | Method and device for zooming pictures | |
| CN107194402B (en) | Parallel refined skeleton extraction method | |
| CN104850232A (en) | Method for acquiring remote gesture tracks under camera conditions | |
| CN104794472A (en) | Method and device used for extracting gesture edge image and gesture extracting method | |
| CN104268845A (en) | Self-adaptive double local reinforcement method of extreme-value temperature difference short wave infrared image | |
| CN103839248A (en) | Image segmentation method and apparatus | |
| JP2018206260A (en) | Image processing system, evaluation model construction method, image processing method, and program | |
| US20210166073A1 (en) | Image generation method and computing device | |
| JP2015186224A (en) | Image processing device, image processing method, and program | |
| CN113077469B (en) | Sketch image semantic segmentation method and device, terminal device and storage medium | |
| Min et al. | Improved range image segmentation by analyzing surface fit patterns | |
| CN113487697A (en) | Method and device for generating simple strokes, electronic equipment and storage medium | |
| JP2000149026A (en) | Image trimming device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| EXSB | Decision made by sipo to initiate substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WD01 | Invention patent application deemed withdrawn after publication | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20150722 |