CN102521567B - Human-computer interaction fingertip detection method, device and television - Google Patents

Human-computer interaction fingertip detection method, device and television Download PDF

Info

Publication number
CN102521567B
CN102521567B CN 201110388105 CN201110388105A CN102521567B CN 102521567 B CN102521567 B CN 102521567B CN 201110388105 CN201110388105 CN 201110388105 CN 201110388105 A CN201110388105 A CN 201110388105A CN 102521567 B CN102521567 B CN 102521567B
Authority
CN
China
Prior art keywords
pixel value
pixel
area
finger tip
circle
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
Application number
CN 201110388105
Other languages
Chinese (zh)
Other versions
CN102521567A (en
Inventor
张登康
邵诗强
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
TCL Corp
Original Assignee
TCL Corp
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by TCL Corp filed Critical TCL Corp
Priority to CN 201110388105 priority Critical patent/CN102521567B/en
Publication of CN102521567A publication Critical patent/CN102521567A/en
Application granted granted Critical
Publication of CN102521567B publication Critical patent/CN102521567B/en
Expired - Fee Related legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Abstract

The invention is applicable to human-computer interaction and provides a human-computer interaction fingertip detection method, a device and a television. The method comprises the following steps of: acquiring an area of interest from inputted video image frames, and extracting binary motion information and binary skin-color information from the area of interest; dividing a palm region according to the binary motion information and the binary skin-color information, and acquiring a binary image containing the palm region; acquiring the central coordinate of the palm region in the binary image, and drawing a circle by taking the central coordinate as the center of the circle and the length of a preset radius of the circle as a radius, wherein the length of the preset radius of the circle is determined according to a preset initial length and a preset iteration step size; and acquiring the pixel values of pixels on a circular path, and carrying out fingertip detection according to the acquired pixel values. With the adoption of the embodiment of the invention, the accuracy in the detection of fingertips can be improved, and the scope of application can be enlarged.

Description

A kind of Fingertip Detection of man-machine interaction, device and televisor
Technical field
The invention belongs to field of human-computer interaction, relate in particular to a kind of Fingertip Detection, device and televisor of man-machine interaction.
Background technology
Along with the development of mode identification technology, increasing man-machine interaction product has appearred on the market, this man-machine interaction product by the identification user finger tip and position, utilize the position of finger tip to realize man-machine interaction, such as icon click, menu affirmation etc.
Existing Fingertip Detection mainly is to carry out finger tip by the mark on data glove or the finger tip to detect, because the method need to increase extra hardware, so high cost, and user's finger also is difficult to flexible motion, makes troubles to the user; And do not carry out the method that finger tip detects by the mark on the finger tip contour accuracy of gesture is not had relatively high expectations, often can only obtain the approximate location of finger tip, if use the method to detect the finger tip of many fingers, the position of then detecting is not accurate enough, the scope of application is less, and is difficult to adapt to various system platforms and occasion.
Summary of the invention
The embodiment of the invention provides a kind of Fingertip Detection of man-machine interaction, is intended to solve existing Fingertip Detection in the inaccurate problem in the existing detection position of finger tip that detects many fingers.
The embodiment of the invention is achieved in that a kind of Fingertip Detection of man-machine interaction, and described method comprises the steps:
Obtain the area-of-interest of the video frame image of input, extract binaryzation movable information and the binaryzation Skin Color Information of described area-of-interest;
Divide palm area according to binaryzation movable information and binaryzation Skin Color Information, obtain the bianry image that comprises described palm area;
Obtain the centre coordinate of palm area in the described bianry image, and take described centre coordinate as the center of circle, default radius of circle length is that radius is justified, described radius of circle length is determined according to default initial length and default iteration step length;
Obtain the pixel value of pixel on the round path, and detect finger tip according to the pixel value that obtains, be specially:
Obtain one by one the pixel value of pixel on the round path;
It is 1 o'clock at the pixel value that obtains, whether the pixel value of judging the left and right sides neighbor of described pixel all is 0, the pixel value that at pixel value is the left and right sides neighbor of 1 pixel all is 0 o'clock, the subregion of judging described pixel place is finger tip, if not, the subregion of judging described pixel place is not finger tip;
Be 0 o'clock at the pixel value that obtains, continue to obtain the pixel value of the next pixel on the round path, until obtain the pixel value of all pixels on the round path;
Wherein, described is that the pixel value of the left and right sides neighbor of 1 pixel is 0 o'clock all at pixel value, and the subregion of judging described pixel place is specially as the step of finger tip:
The pixel value that at pixel value is the left and right sides neighbor of 1 pixel all is 0 o'clock, and centered by 1 pixel, default expansion multiple is expanded, and the zone after will expanding is divided into 4 zones take pixel value;
Detect the pixel value of pixel in 4 zones, if the pixel value of pixel all is 1 in one of them zone, and the pixel value of all the other regional pixels is 0, judges that then the central area in the zone after expanding is finger tip.
Another purpose of the embodiment of the invention is to provide a kind of finger tip pick-up unit of man-machine interaction, and described device comprises:
The area-of-interest acquiring unit for the area-of-interest of the video frame image that obtains input, extracts binaryzation movable information and the binaryzation Skin Color Information of described area-of-interest;
The palm area acquiring unit is used for dividing palm area according to binaryzation movable information and binaryzation Skin Color Information, obtains the bianry image that comprises described palm area;
Circle path acquiring unit, be used for obtaining the centre coordinate of described bianry image palm area, and take described centre coordinate as the center of circle, default radius of circle length is that radius is justified, and described radius of circle length is determined according to default initial length and default iteration step length;
The finger tip detecting unit, for the pixel value that obtains pixel on the round path, and according to the pixel value detection finger tip that obtains;
Described finger tip detecting unit comprises:
The pixel value acquisition module is for the pixel value that obtains one by one pixel on the round path;
The first finger tip judge module, being used at the pixel value that obtains is 1 o'clock, whether the pixel value of judging the left and right sides neighbor of described pixel all is 0, the pixel value that at pixel value is the left and right sides neighbor of 1 pixel all is 0 o'clock, the subregion of judging described pixel place is finger tip, if not, the subregion of judging described pixel place is not finger tip;
Trigger module, being used at the pixel value that obtains is 0 o'clock, triggers described pixel value acquisition module and continues to obtain the pixel value of the next pixel on the round path, until obtain the pixel value of all pixels on the round path;
Wherein, described the first finger tip judge module specifically is used for:
The pixel value that at pixel value is the left and right sides neighbor of 1 pixel all is 0 o'clock, and centered by 1 pixel, default expansion multiple is expanded, and the zone after will expanding is divided into 4 zones take pixel value;
Detect the pixel value of pixel in 4 zones, if the pixel value of pixel all is 1 in one of them zone, and the pixel value of all the other regional pixels is 0, judges that then the central area in the zone after expanding is finger tip.
Another purpose of the embodiment of the invention is to provide a kind of televisor that comprises the finger tip pick-up unit of above-mentioned man-machine interaction.
The embodiment of the invention is by obtaining the area-of-interest of video image series, and obtain binaryzation movable information and binaryzation Skin Color Information according to this area-of-interest, extract the bianry image that comprises palm area according to the binaryzation movable information that obtains and binaryzation Skin Color Information, detect again user's finger tip from the bianry image that comprises palm area that extracts.Because the embodiment of the invention is by dividing the bianry image that palm area is extracted palm area in conjunction with binaryzation movable information and binaryzation Skin Color Information, therefore can effectively reject the interference of class area of skin color in the background, make the palm area of extraction more accurate, thereby improve the accuracy that checks finger tip, and in the process that detects finger tip, need not Fingers to directly over, also can correctly detect finger tip, enlarged range of application.
Description of drawings
Fig. 1 is the process flow diagram of the Fingertip Detection of the man-machine interaction that provides of first embodiment of the invention;
Fig. 2 is the area-of-interest figure that determines in the Fingertip Detection of the man-machine interaction that provides of second embodiment of the invention;
Fig. 3 is the binaryzation movable information figure that is converted to according to Fig. 2 in the Fingertip Detection of the man-machine interaction that provides of second embodiment of the invention;
Fig. 4 is the binaryzation Skin Color Information figure that is converted to according to the Skin Color Information of the HSV color space of Fig. 2 in the Fingertip Detection of the man-machine interaction that provides of second embodiment of the invention;
Fig. 5 comprises the palm area binary map according to what Fig. 3 and Fig. 4 obtained in the Fingertip Detection of the man-machine interaction that provides of third embodiment of the invention;
Fig. 6 is that a plurality of subregions of in the Fingertip Detection of the man-machine interaction that provides of third embodiment of the invention Fig. 5 being divided re-execute the binary map that obtains after the binary conversion treatment;
Round path binary map has been provided on the basis of Fig. 6 in the Fingertip Detection of the man-machine interaction that provides of fourth embodiment of the invention Fig. 7;
Fig. 8 is the process flow diagram of the Fingertip Detection of the man-machine interaction that provides of fifth embodiment of the invention;
Fig. 9 is the structural representation of the finger tip pick-up unit of the man-machine interaction that provides of sixth embodiment of the invention;
Figure 10 is the structural representation of the finger tip pick-up unit of the man-machine interaction that provides of seventh embodiment of the invention;
Figure 11 is the structural representation of the finger tip pick-up unit of the man-machine interaction that provides of eighth embodiment of the invention;
Figure 12 is the structural representation of the finger tip pick-up unit of the man-machine interaction that provides of ninth embodiment of the invention;
Figure 13 is the structural representation of the finger tip pick-up unit of the man-machine interaction that provides of tenth embodiment of the invention;
Figure 14 is the structural representation of the finger tip pick-up unit of the man-machine interaction that provides of eleventh embodiment of the invention.
Embodiment
In order to make purpose of the present invention, technical scheme and advantage clearer, below in conjunction with drawings and Examples, the present invention is further elaborated.Should be appreciated that specific embodiment described herein only in order to explain the present invention, is not intended to limit the present invention.
The embodiment of the invention according to the region of interesting extraction movable information and the Skin Color Information that obtain, merges this movable information and Skin Color Information to obtain the bianry image that comprises hand region, detects the finger tip of finger from the bianry image that comprises hand region again.
It is a kind of that the embodiment of the invention provides: the Fingertip Detection of man-machine interaction, device and televisor.
Described method comprises: obtain the area-of-interest of the video frame image of input, extract binaryzation movable information and the binaryzation Skin Color Information of described area-of-interest;
Divide palm area according to binaryzation movable information and binaryzation Skin Color Information, obtain the bianry image that comprises described palm area;
Obtain the centre coordinate of palm area in the described bianry image, and take described centre coordinate as the center of circle, default radius of circle length is that radius is justified, described radius of circle length is determined according to default initial length and default iteration step length;
Obtain the pixel value of pixel on the round path, and detect finger tip according to the pixel value that obtains, be specially:
Obtain one by one the pixel value of pixel on the round path;
It is 1 o'clock at the pixel value that obtains, whether the pixel value of judging the left and right sides neighbor of described pixel all is 0, the pixel value that at pixel value is the left and right sides neighbor of 1 pixel all is 0 o'clock, the subregion of judging described pixel place is finger tip, if not, the subregion of judging described pixel place is not finger tip;
Be 0 o'clock at the pixel value that obtains, continue to obtain the pixel value of the next pixel on the round path, until obtain the pixel value of all pixels on the round path;
Wherein, described is that the pixel value of the left and right sides neighbor of 1 pixel is 0 o'clock all at pixel value, and the subregion of judging described pixel place is specially as the step of finger tip:
The pixel value that at pixel value is the left and right sides neighbor of 1 pixel all is 0 o'clock, and centered by 1 pixel, default expansion multiple is expanded, and the zone after will expanding is divided into 4 zones take pixel value;
Detect the pixel value of pixel in 4 zones, if the pixel value of pixel all is 1 in one of them zone, and the pixel value of all the other regional pixels is 0, judges that then the central area in the zone after expanding is finger tip.
Described device comprises: the area-of-interest acquiring unit, for the area-of-interest of the video frame image that obtains input, extract binaryzation movable information and the binaryzation Skin Color Information of described area-of-interest;
The palm area acquiring unit is used for dividing palm area according to binaryzation movable information and binaryzation Skin Color Information, obtains the bianry image that comprises described palm area;
Circle path acquiring unit, be used for obtaining the centre coordinate of described bianry image palm area, and take described centre coordinate as the center of circle, default radius of circle length is that radius is justified, and described radius of circle length is determined according to default initial length and default iteration step length;
The finger tip detecting unit, for the pixel value that obtains pixel on the round path, and according to the pixel value detection finger tip that obtains;
Described finger tip detecting unit comprises:
The pixel value acquisition module is for the pixel value that obtains one by one pixel on the round path;
The first finger tip judge module, being used at the pixel value that obtains is 1 o'clock, whether the pixel value of judging the left and right sides neighbor of described pixel all is 0, the pixel value that at pixel value is the left and right sides neighbor of 1 pixel all is 0 o'clock, the subregion of judging described pixel place is finger tip, if not, the subregion of judging described pixel place is not finger tip;
Trigger module, being used at the pixel value that obtains is 0 o'clock, triggers described pixel value acquisition module and continues to obtain the pixel value of the next pixel on the round path, until obtain the pixel value of all pixels on the round path;
Wherein, described the first finger tip judge module specifically is used for:
The pixel value that at pixel value is the left and right sides neighbor of 1 pixel all is 0 o'clock, and centered by 1 pixel, default expansion multiple is expanded, and the zone after will expanding is divided into 4 zones take pixel value;
Detect the pixel value of pixel in 4 zones, if the pixel value of pixel all is 1 in one of them zone, and the pixel value of all the other regional pixels is 0, judges that then the central area in the zone after expanding is finger tip.
Described equipment comprises: a kind of televisor that comprises the finger tip pick-up unit of above-mentioned man-machine interaction.
The embodiment of the invention is by obtaining the area-of-interest of video image series, and obtain binaryzation movable information and binaryzation Skin Color Information according to this area-of-interest, extract the bianry image that comprises palm area according to the binaryzation movable information that obtains and binaryzation Skin Color Information, detect again user's finger tip from the bianry image that comprises palm area that extracts.Because the embodiment of the invention is by dividing the bianry image that palm area is extracted palm area in conjunction with binaryzation movable information and binaryzation Skin Color Information, therefore can effectively reject the interference of class area of skin color in the background, make the palm area of extraction more accurate, thereby improve the accuracy that checks finger tip, and in the process that detects finger tip, need not Fingers to directly over, also can correctly detect finger tip, enlarged range of application.
For technical solutions according to the invention are described, describe below by specific embodiment.
Embodiment one:
Fig. 1 shows the process flow diagram of the Fingertip Detection of the man-machine interaction that first embodiment of the invention provides, in the present embodiment, according to the region of interesting extraction movable information and the Skin Color Information that obtain, merge this movable information and Skin Color Information to obtain the bianry image that comprises hand region, from the bianry image that comprises hand region, detect again the finger tip of finger.Details are as follows:
Step S11 obtains the area-of-interest of the video frame image of input, extracts binaryzation movable information and the binaryzation Skin Color Information of this area-of-interest.
In the embodiment of the invention, the video frame image of input is mainly coloured image, after the area-of-interest that has obtained video frame image, according to the motion feature of area-of-interest, binaryzation movable information and the binaryzation Skin Color Information that features of skin colors extracts respectively this area-of-interest.
Step S12 divides palm area according to binaryzation movable information and binaryzation Skin Color Information, obtains the bianry image that comprises this palm area.
In the present embodiment, the palm positional information and the binaryzation Skin Color Information that utilize the binaryzation movable information to obtain are separated with palm area, obtain comprising two-value (0, the 1) image of this palm area.
Step S13 obtain the centre coordinate of palm area in the bianry image, and take this centre coordinate as the center of circle, default radius of circle length is that radius is justified, and this radius of circle length is determined according to default initial length and default iteration step length.
Step S14 obtains the pixel value of pixel on the round path, and detects finger tip according to the pixel value that obtains.
In the present embodiment, determine default radius of circle according to default initial length and iteration step length.Justify as the center of circle take this default radius of circle and the centre coordinate of palm area, and obtain the pixel value of the pixel on the round round path of doing, judge according to the pixel value that obtains whether position corresponding to this pixel value is finger tip again.
In first embodiment of the invention, obtain the area-of-interest of video image series, and obtain binaryzation movable information and binaryzation Skin Color Information according to this area-of-interest, extract the bianry image that comprises palm area according to the binaryzation movable information that obtains and binaryzation Skin Color Information, detect again user's finger tip from the bianry image that comprises palm area that extracts.Because the embodiment of the invention is by dividing the bianry image that palm area is extracted palm area in conjunction with binaryzation movable information and binaryzation Skin Color Information, therefore can effectively reject the interference of class area of skin color in the background, make the palm area of extraction more accurate, thereby improve the accuracy that detects finger tip, and in the process that detects finger tip, need not Fingers to directly over, also can correctly detect finger tip, enlarged range of application.
Embodiment two:
Second embodiment of the invention is mainly done more at large to describe to the step S11 of embodiment one, and all the other steps are identical with embodiment's one, repeat no more herein.
Wherein step S11 is specially:
A1, determine the area-of-interest of video frame image.
In the present embodiment, video image series to input delimited an area-of-interest, such as, when the color video frequency image series of input comprises all features of people, the situation of carrying out the corresponding command according to people's fingertip motions according to man-machine interaction, the area-of-interest of then delimiting is generally that to have got rid of the Skin Color Information zone different from the palm Skin Color Information and Skin Color Information identical with the palm Skin Color Information but be static zone, such as, the area-of-interest of delimiting can be for comprising the whole arm regions of motion finger tip, this arm regions has been got rid of human face region and another arm regions, specifically as shown in Figure 2.
A2, the area-of-interest of video frame image is converted to gray level image, and adopts the three-frame difference method to detect the motion target area of gray level image, extract the binaryzation movable information according to this motion target area again.
In the present embodiment, area-of-interest to video frame image is converted to gray level image, all continuous three frame video images that recycle this gray level image carry out the three-frame difference method, to detect the region of variation of moving target itself, the greyscale image transitions that will detect at last the region of variation of moving target itself is bianry image, and from this bianry image extraction binaryzation movable information, this binaryzation movable information comprises the positional information of palm area etc.Wherein, the binaryzation movable information image that is converted to according to the region of variation of moving target itself of area-of-interest specifically as shown in Figure 3.
A3, the default conversion formula of basis are converted to hue saturation value HSV color space with the area-of-interest that obtains from the RGB rgb color space, extract the Skin Color Information of this HSV color space, extract the binaryzation Skin Color Information according to this Skin Color Information again.
In the present embodiment, HSV color space model is by tone (Hue, H), color saturation (Saturation, S), and the base attribute of 3 colors of brightness value (Value, V) color of object is described, H represents the position of spectral color of living in, and scope is [0,360]; S is a ratio value, and scope is [0,1], and it represents the ratio between the purity of the purity of selected color and this color maximum; V represents the bright degree of color, and scope is [0,1].The HSV color space belongs to non-linear color representation space, and its saturation degree and colourity can reflect the color attributes of target accurately, and consistent to the perception of color with human body.According to default conversion formula, the video image of inputting is converted to the HSV color space from rgb color space, the formula of conversion is specially:
Figure GDA00003497748700091
Figure GDA00003497748700092
V＝max(r,g,b)
Wherein, r, g, b are respectively the value of a pixel on R, G, B component.
Further, mainly concentrate on this feature of red area according to the skin tone of human body, there is following relation: r in this feature in image〉g〉b, accordingly above-mentioned conversion formula is simplified to reduce calculated amount, the formula of wherein simplifying is:
S = r - b r
H = ( b - g ) * π / 3 r - b
V＝r
After the Skin Color Information that has obtained the HSV color space, this Skin Color Information is converted to bianry image information, to extract the binaryzation Skin Color Information, wherein, according to the binaryzation Skin Color Information image of the Skin Color Information conversion of area-of-interest specifically as shown in Figure 4.
In the second embodiment of the invention, at first obtain the area-of-interest of inputted video image frame, and obtain binaryzation movable information and binaryzation Skin Color Information from this area-of-interest.Obtain the binaryzation movable information owing to adopt the three-frame difference method, adopt default conversion formula to obtain the binaryzation Skin Color Information, therefore can guarantee that two value informations that obtain are more accurate, so follow-up according to binaryzation Skin Color Information and binaryzation movable information division palm area.
Embodiment three:
Third embodiment of the invention is mainly done more at large to describe to the step S12 of embodiment one, and all the other steps are identical with embodiment one or embodiment's two, repeat no more herein.
Wherein step S12 is specially:
B1, divide palm area according to binaryzation movable information and binaryzation Skin Color Information, obtain bianry image corresponding to palm area.
In the present embodiment, determine palm area in conjunction with the binaryzation movable information that obtains and binaryzation Skin Color Information, improve the accuracy of the palm area of determining.
Wherein, bianry image corresponding to the palm area of obtaining in Fig. 5, uses white expression palm area as shown in Figure 5, use black to represent other zones, therefore show that the pixel value in the palm area is 1, other regional pixel values are 0, certainly, also can use black to represent palm area, use other zones of white expression, be not construed as limiting herein, present embodiment uses white expression palm area.
B2, bianry image corresponding to palm area of dividing is divided into a plurality of subregions, and a plurality of subregions of dividing carried out morphology process.
In the present embodiment, because the existence of noise, a plurality of isolated zonules and little gap often appear in the bianry image that therefore comprises palm area, it is exactly to eliminate the isolated zonule that occurs and the little gap of filling appearance that this bianry image that comprises palm area is carried out the morphology processing, and the morphology processing of eliminating the isolated zonule that occurs is called corrosion, the closely spaced morphology that fill to occur is processed and is called expansion, and the processing procedure of dilation and erosion is the process of the part in image or the image and predefined nuclear being carried out convolution algorithm.
Wherein, bianry image corresponding to palm area of dividing is divided into a plurality of subregions, and the step that a plurality of subregions execution morphology of dividing are processed is specially:
C1, according to the needs of actual conditions, the bianry image that comprises palm area that obtains is divided into default a plurality of subregions, and a plurality of subregions of dividing is re-executed binary conversion treatment.In the present embodiment, as required, the bianry image that comprises palm area that obtains is divided into a plurality of subregions, with this palm area gridding.Such as, be divided into N * N sub regions, then the size of every sub regions is: the length/N of the length=palm area of subregion; Subregion wide=palm area wide/N.Default number of pixels threshold value, judge that pixel value in the subregion is whether 1 number of pixels is greater than default number of pixels threshold value, when pixel value is 1 number of pixels greater than default number of pixels threshold value in subregion, this subregion is judged to be a pixel, and the pixel value that defines this pixel is 1, otherwise the pixel value that defines this pixel is 0.In bianry image, pixel value is that 1 zone is white, and pixel value is that 0 zone is black.Wherein, Fig. 6 re-executes the binary map that obtains after the binary conversion treatment for a plurality of subregions that the bianry image of Fig. 5 is divided.
C2, adopt predefined nuclear to eliminate the soliton zone of palm area, perhaps fill the little gap of palm area, the soliton zone of this palm area and the little gap of palm area are the subregion that re-executes after the binary conversion treatment.In the present embodiment, the nuclear of the dilation and erosion during predefine morphology is processed, generally, the nuclear that expands and the nuclear of corrosion are that a centre is with filled squares or the disk of reference point, the nuclear size of this expansion and the nuclear size of corrosion are greater than all subregion of dividing, when all subregion was regarded as a pixel, the nuclear size of the nuclear of this expansion size and corrosion was greater than a pixel, otherwise was difficult to eliminate preferably or fill all subregion of division.
Present embodiment take reference point at a RC foursquare nuclear as example, all subregion is expanded or corrosion treatment.Wherein, the step of all subregion being carried out expansion process is: adopt the canny operator to detect the profile of palm, the largest connected territory that obtains palm judges in the largest connected territory of this palm whether have the gap, if exist, then uses predefined nuclear to fill this gap.In the present embodiment, the gap that exists in the largest connected territory of palm is the pixel value zone different from other area pixel values, and when being white such as the most of zone at palm, the gap in the palm is the zone of black.Use predefined nuclear to fill this black region, then the pixel value in gap is become 1 from 0, thereby make the pixel value in gap identical with pixel value in the palm area, realized the filling in gap.The step of all subregion being carried out corrosion treatment is: judge whether the largest connected overseas of palm exists the soliton zone, if exist, then use predefined nuclear to eliminate this soliton zone.In the present embodiment, the zone of soliton zone for protruding outside the palm border.
In the third embodiment of the invention, process carrying out morphology according to the palm area of binaryzation movable information and the division of binaryzation Skin Color Information, this morphology is processed by the cavity of the palm area of the palm binary image being carried out grid and divide, fill up division or is eliminated the soliton zone that the palm area of dividing is protruded, optimized the palm profile, strengthened the characteristic of unique point on the profile, particularly strengthened the unique point of fingertip area, make fingertip area more obvious, thereby improved accuracy, the continuity of dividing palm area, and the accuracy rate of finger tip detection.
Embodiment four:
Fourth embodiment of the invention is mainly done more at large to describe to step S13 and the step S14 of embodiment one, and all the other steps are identical with embodiment one or embodiment two or embodiment's three, repeat no more herein.
Step S13 wherein obtain the centre coordinate of palm area in the bianry image, and take this centre coordinate as the center of circle, default radius of circle length is that radius is justified, and this radius of circle length is determined to be specially according to default initial length and default iteration step length:
D1, determine the circle centre coordinate: the pixel value of point by point scanning palm area, determine the centre coordinate of palm area according to pixel value and pixel value position.
In the present embodiment, obtain coordinate and the pixel value corresponding to this pixel of pixel in the palm area, and the centre coordinate of determining palm area according to coordinate and the pixel value corresponding to this pixel of the pixel of obtaining, such as, determine the centre coordinate (x, y) of palm area according to following formula:
x = Σ I ( i , j ) * i Σi
y = Σ I ( i , j ) * j Σj
Wherein, I (i, j)For coordinate position in the image is the pixel value of (i, j), x and y are respectively x component and the y component of the centre coordinate of palm area.
D2, determine radius of circle length: the radius of circle length of present embodiment can or be subtracted each other definite by default initial length and iteration step length by default initial length and iteration step length addition.
Further, default initial length can be the minimum rectangular length in the largest connected territory that comprises palm area.
D3, justify according to circle centre coordinate and radius of circle length: in the present embodiment, take the centre coordinate of palm area as the center of circle, default radius of circle length is that radius is justified, and specifically as shown in Figure 7, the center of circle O of the circle among this Fig. 7 is the centre coordinate of the palm area of Fig. 6.
Step S14 wherein obtains the pixel value of pixel on the round path, and detects finger tip according to the pixel value that obtains and be specially:
In the present embodiment, after obtaining circle, obtain the pixel value of pixel on the round path, and detect finger tip according to the pixel value that obtains.
E1, obtain the pixel value of pixel on the round path one by one;
E2, it is 1 o'clock at the pixel value that obtains, whether the pixel value of judging the left and right sides neighbor of this pixel all is 0, the pixel value that at pixel value is the left and right sides neighbor of 1 pixel all is 0 o'clock (zone on every side all is blank), the subregion of judging this pixel place is finger tip, if not, the subregion of judging this pixel place is not finger tip;
E3, be 0 o'clock at the pixel value that obtains, continue to obtain the pixel value of the next pixel on the round path, until obtain the pixel value of all pixels on the round path.
Certainly, present embodiment is to detect finger tip with 1 expression palm area, also can detect finger tip with 0 expression palm area, and concrete detecting step is similar to the above, repeats no more herein.
Further, be that the pixel value of the left and right sides neighbor of 1 pixel is 0 o'clock all at pixel value, the subregion of judging this pixel place is specially as the step of finger tip:
The pixel value that at pixel value is the left and right sides neighbor of 1 pixel all is 0 o'clock, and centered by 1 pixel, default expansion multiple is expanded, and the zone after will expanding is divided into 4 zones take pixel value; Detect the pixel value of pixel in 4 zones, if the pixel value of pixel all is 1 in one of them zone, and the pixel value of all the other regional pixels is 0, judges that then the central area in the zone after expanding is finger tip.For example, default expansion multiple is 9, and then the area size after the expansion becomes 9 * 9 zone, and the pixel value that is in the pixel of 9 * 9 regional centers is 1, satisfies above-mentioned condition such as this pixel, judges that then zone corresponding to this pixel is finger tip.
In the present embodiment, determine whether finger tip by enlarging surveyed area, further improved the accurate rate that finger tip detects.
Embodiment five:
Fig. 8 shows the Fingertip Detection flow process of the man-machine interaction that fifth embodiment of the invention provides, and adopts the method for present embodiment can detect all finger tips that the user carries out man-machine interaction, and details are as follows:
Wherein the step S11 with embodiment one is identical with step S12 respectively with step S22 for step S21, repeats no more herein.
Step S23 obtain the centre coordinate of palm area in the bianry image, and take this centre coordinate as the center of circle, default initial length is that radius is justified.
In the present embodiment, default initial length can be the minimum rectangular length in the largest connected territory that comprises palm area, also can be half of the minimum rectangular length in the largest connected territory that comprises palm area, is not construed as limiting herein.
Step S24 obtains the pixel value of pixel on the round path, and detects finger tip according to the pixel value that obtains.
In the present embodiment, concrete Fingertip Detection is identical with embodiment's three, repeats no more herein.
Step S25 determines next radius of circle according to default iteration step length and default initial length, the difference of judging the length of this next one radius of circle and preset length threshold value whether in preset range, if, execution in step S26, otherwise, execution in step S27.
In the present embodiment, suppose that default initial length is the minimum rectangular length that comprises the largest connected territory of palm area, then initial length that should be default and default iteration step length can be subtracted each other to determine next radius of circle, and the difference of judging the length of this next one radius of circle and preset length threshold value is whether in preset range.Wherein, the preset length threshold value is for more than or equal to 0 and less than or equal to the number of the minimum rectangular length in the largest connected territory that comprises palm area.
Certainly, when default initial length was worth for other, radius of circle also can be determined by default initial length and default iteration step length addition, repeat no more herein.
Step S26, take the centre coordinate of palm area as the center of circle, next radius of circle is that radius is justified, and forwards step S24 to.
In the present embodiment, determine that according to step S25 after the next radius of circle, take this next one radius of circle as radius, the centre coordinate of palm area is that justify in the center of circle.
Step S27 stops finger tip and detects.
In the present embodiment, if detected the scope not far apart from the center of palm area, such as detecting apart from the center of the palm area length less than 1/4th the largest connected territory that comprises palm area, then stop to detect finger tip.
In embodiments of the present invention, iteration changes radius of circle length again justify, and according to the pixel detection finger tip on the round path of the circle of again doing, the execution above-mentioned steps that circulates, thus can accurately detect a plurality of finger tips.
Further, after the step that stops the finger tip detection, further comprise step S28:
Step S28 rejects the unreasonable finger tip that has been judged to be finger tip.
Wherein, the step of rejecting the unreasonable finger tip be judged to be finger tip is specially:
F1, judge the subregion be judged to be finger tip and the left and right sides adjacent area of this subregion distance whether all greater than the distance threshold of presetting, wherein, the left and right sides adjacent area of this subregion and the pixel value of the pixel between this subregion are 0.
F2, during all greater than default distance threshold, reject the finger tip that this has been judged to be finger tip in the distance of the left and right sides of the subregion that is judged to be finger tip and this subregion adjacent area; Otherwise, keep the finger tip that has been judged to be finger tip.
In the present embodiment, the zone that is judged to be finger tip is judged again, rejected the zone that is mistaken for finger tip, so that the finger tip that keeps is more accurate, improve the accuracy rate that finger tip detects.
Embodiment six:
Fig. 9 shows the structure of the finger tip pick-up unit of the man-machine interaction that sixth embodiment of the invention provides, and for convenience of explanation, only shows the part relevant with the embodiment of the invention.
The finger tip pick-up unit of this man-machine interaction can be used for the various information processing terminals by wired or wireless network connection server, pocket computing machine (Pocket Personal Computer for example, PPC), computing machine, televisor, notebook computer etc., can be to run on the unit that software unit, hardware cell or software and hardware in these terminals combine, also can be used as independently, suspension member is integrated in these terminals or runs in the application system of these terminals, wherein:
Area-of-interest acquiring unit 81 for the area-of-interest of the video frame image that obtains input, extracts binaryzation movable information and the binaryzation Skin Color Information of this area-of-interest.
Palm area acquiring unit 82 is used for dividing palm area according to binaryzation movable information and binaryzation Skin Color Information, obtains the bianry image that comprises this palm area.
Circle path acquiring unit 83, for the centre coordinate that obtains this bianry image palm area, and take this centre coordinate as the center of circle, default radius of circle length is that radius is justified, this radius of circle length is definite with default iteration step length according to default initial length.
Finger tip detecting unit 84, for the pixel value that obtains pixel on the round path, and according to the pixel value detection finger tip that obtains.
The finger tip pick-up unit of the man-machine interaction that the embodiment of the invention six provides can use in the Fingertip Detection that previous embodiment one provides, and details are referring to the description of above-described embodiment one.
In sixth embodiment of the invention, obtain the area-of-interest of video image series, and obtain binaryzation movable information and binaryzation Skin Color Information according to this area-of-interest, extract the bianry image that comprises palm area according to the binaryzation movable information that obtains and binaryzation Skin Color Information, detect again user's finger tip from the bianry image that comprises palm area that extracts.Because the embodiment of the invention is by dividing the bianry image that palm area is extracted palm area in conjunction with binaryzation movable information and binaryzation Skin Color Information, therefore can effectively reject the interference of class area of skin color in the background, make the palm area of extraction more accurate, thereby improve the accuracy that checks finger tip, and in the process that detects finger tip, need not Fingers to directly over, also can correctly detect finger tip, enlarged range of application.
Embodiment seven:
Figure 10 shows the structural representation of the finger tip pick-up unit of the man-machine interaction that seventh embodiment of the invention provides, the main refinement of present embodiment area-of-interest acquiring unit 81, remaining element is identical with embodiment five, repeats no more herein:
This area-of-interest acquiring unit 81 comprises: area-of-interest delimited module 811, extraction of motion information module 812 and Skin Color Information extraction module 813.
This area-of-interest delimited the area-of-interest that module 811 is used for determining video frame image.
Extraction of motion information module 812 is used for the area-of-interest of video frame image is converted to gray level image, and adopts the three-frame difference method to detect the motion target area of gray level image, extracts the binaryzation movable information according to this motion target area again.
Skin Color Information extraction module 813 is converted to hue saturation value HSV color space with the area-of-interest that obtains from the RGB rgb color space according to default conversion formula, extract the Skin Color Information of this HSV color space, extract the binaryzation Skin Color Information according to this Skin Color Information again.
In the present embodiment, default conversion formula is as follows:
Figure GDA00003497748700172
V＝max(r,g,b)
Wherein, r, g, b are respectively the value of a pixel on R, G, B component.
Further, mainly concentrate on this feature of red area according to the skin tone of human body, there is following relation: r in this feature in image〉g〉b, accordingly above-mentioned conversion formula is reduced to:
S = r - b r
H = ( b - g ) * π / 3 r - b
V＝r
The finger tip pick-up unit of the man-machine interaction that the embodiment of the invention seven provides can use in the Fingertip Detection that previous embodiment two provides, and details are referring to the description of above-described embodiment two.
In the seventh embodiment of the invention, owing to adopt the three-frame difference method to obtain the binaryzation movable information, adopt default conversion formula to obtain the binaryzation Skin Color Information, therefore can guarantee that two value informations that obtain are more accurate.
Embodiment eight:
Figure 11 shows the structural representation of the finger tip pick-up unit of the man-machine interaction that eighth embodiment of the invention provides, the main refinement of present embodiment palm area acquiring unit 82, remaining element is identical with embodiment five or embodiment six, repeats no more herein:
This palm area acquiring unit 82 comprises: two-value palm area determination module 821 and denoising module 822.
Two-value palm area determination module 821 is used for dividing palm area according to binaryzation movable information and binaryzation Skin Color Information.
Denoising module 822, bianry image corresponding to palm area that is used for dividing is divided into a plurality of subregions, and a plurality of subregions of dividing carried out morphology process.
The finger tip pick-up unit of the man-machine interaction that the embodiment of the invention eight provides can use in the Fingertip Detection that previous embodiment three provides, and details are referring to the description of above-described embodiment three.
In the present embodiment, because the existence of noise, a plurality of isolated zonules and little gap often appear in the bianry image that therefore comprises palm area, and it is exactly to eliminate the isolated zonule that occurs and the little gap of filling appearance that this bianry image that comprises palm area is carried out the morphology processing.At first according to the needs of actual conditions, the bianry image that comprises palm area that obtains is divided into default a plurality of subregions, and a plurality of subregions of dividing are re-executed binary conversion treatment; Adopt predefined nuclear to eliminate the soliton zone of palm area again, perhaps fill the little gap of palm area, the soliton zone of this palm area and the little gap of palm area are the subregion that re-executes after the binary conversion treatment.Generally, the nuclear of expansion and the nuclear of corrosion be a centre with filled squares or the disk of reference point, the nuclear size of the nuclear of this expansion size and corrosion is greater than all subregion of dividing.
Embodiment nine:
Figure 12 shows the structural representation of the finger tip pick-up unit of the man-machine interaction that ninth embodiment of the invention provides, the main refinement of present embodiment round path acquiring unit 83, remaining element and embodiment five or embodiment six or embodiment seven are identical, repeat no more herein:
This circle path acquiring unit 83 comprises: radius of circle determination module 831, center of circle determination module 832, circle path determination module 833.
Radius of circle determination module 831 is used for determining radius of circle length by default initial length and iteration step length addition, or subtracts each other definite radius of circle length by default initial length and iteration step length.
In the present embodiment, determine radius of circle length according to default initial length and iteration step length.Wherein, default initial length can be made as the minimum rectangular length in the largest connected territory that comprises palm area.
Center of circle determination module 832 is used for the pixel value of point by point scanning palm area, determines the centre coordinate of palm area according to the pixel value that obtains and pixel value position.
In the present embodiment, obtain coordinate and the pixel value corresponding to this pixel of pixel in the palm area, and the centre coordinate of determining palm area according to coordinate and the pixel value corresponding to this pixel of the pixel of obtaining, such as, determine the centre coordinate (x, y) of palm area according to following formula:
x = Σ I ( i , j ) * i Σi
y = Σ I ( i , j ) * j Σj
Wherein, I (i, j)For coordinate position in the image is the pixel value of (i, j), x and y are respectively x component and the y component of the centre coordinate of palm area.
Circle path determination module 833 is used for centre coordinate take palm area as the center of circle, justifies take this radius of circle length of determining as radius.
The finger tip pick-up unit of the man-machine interaction that the embodiment of the invention nine provides can use in the Fingertip Detection that previous embodiment four provides, and details are referring to the description of above-described embodiment four.
In the ninth embodiment of the invention, go out radius of circle length according to default initial length and iteration step length iteration, centered by this radius of circle length and palm area, justify, so that the subsequent detection finger tip.
Embodiment ten:
Figure 13 shows the structural representation of the finger tip pick-up unit of the man-machine interaction that tenth embodiment of the invention provides, the main refinement of present embodiment finger tip detecting unit 84, remaining element and embodiment five or embodiment six or embodiment seven or embodiment eight are identical, repeat no more herein:
This finger tip detecting unit 84 comprises pixel value acquisition module 841, the first finger tip judge module 842, the second finger tip judge module 843.
Pixel value acquisition module 841 is used for obtaining one by one the pixel value of pixel on the round path.
It is 1 o'clock that the first finger tip judge module 842 is used at the pixel value that obtains, whether the pixel value of judging the left and right sides neighbor of this pixel all is 0, the pixel value that at pixel value is the left and right sides neighbor of 1 pixel is 0 o'clock all, judges that the subregion at this pixel place is finger tip.
In the present embodiment, be that the pixel value of the left and right sides neighbor of 1 pixel all is 0 o'clock at pixel value, centered by 1 pixel, default expansion multiple is expanded, and the zone after will expanding is divided into 4 zones take pixel value; Detect the pixel value of pixel in 4 zones, if the pixel value of pixel all is 1 in one of them zone, and the pixel value of all the other regional pixels is 0, judges that then the central area in the zone after expanding is finger tip.For example, suppose that default expansion multiple is 9, then the area size after the expansion becomes 9 * 9 zone, and the pixel value that is in the pixel of 9 * 9 regional centers is 1, satisfies above-mentioned condition such as this pixel, judges that then zone corresponding to this pixel is finger tip.
It is 0 o'clock that trigger module 843 is used at the pixel value that obtains, and triggers described pixel value acquisition module 841 and continues to obtain the pixel value of the next pixel on the round path, until obtain the pixel value of all pixels on the round path.
Certainly, present embodiment is to detect finger tip with 1 expression palm area, also can detect finger tip with 0 expression palm area, and concrete detecting step is similar to the above, repeats no more herein.
In the present embodiment, by the expansion surveyed area, thereby improved the accurate rate that finger tip detects.The finger tip pick-up unit of the man-machine interaction that the embodiment of the invention ten provides can use in the Fingertip Detection that previous embodiment four provides, and details are referring to the description of above-described embodiment four.
Embodiment 11:
Figure 14 shows the structural representation of the finger tip pick-up unit of the man-machine interaction that eleventh embodiment of the invention provides, and for convenience of explanation, only shows the part relevant with present embodiment:
In the embodiment of the invention, the finger tip pick-up unit of man-machine interaction also comprises: region distance judging unit 85 and finger tip culling unit 86, other unit of the finger tip pick-up unit of this man-machine interaction are identical with embodiment five or embodiment six or embodiment seven or embodiment eight or embodiment's nine, repeat no more herein.
Whether region distance judging unit 85 is used for judging the distance of the subregion that has been judged to be finger tip and the left and right sides adjacent area of this subregion all greater than the distance threshold of presetting, and the left and right sides adjacent area of this subregion and the pixel value of the pixel between this subregion are 0.
Finger tip culling unit 86 is used for during all greater than the distance threshold preset, rejecting the finger tip that this has been judged to be finger tip in the distance of the left and right sides of the subregion that is judged to be finger tip and this subregion adjacent area.
In eleventh embodiment of the invention, the zone that is judged to be finger tip is judged again, rejected the zone that is mistaken for finger tip, so that the finger tip that keeps is more accurate, improve the accuracy rate that finger tip detects.
The finger tip pick-up unit of the man-machine interaction that the embodiment of the invention 11 provides can use in the Fingertip Detection that previous embodiment five provides, and details are referring to the description of above-described embodiment five.
The finger tip pick-up unit of the man-machine interaction that the embodiment of the invention provides is by obtaining the area-of-interest of video image series, and obtain binaryzation movable information and binaryzation Skin Color Information according to this area-of-interest, extract the bianry image that comprises palm area according to the binaryzation movable information that obtains and binaryzation Skin Color Information, detect again user's finger tip from the bianry image that comprises palm area that extracts.Because the embodiment of the invention is by dividing the bianry image that palm area is extracted palm area in conjunction with binaryzation movable information and binaryzation Skin Color Information, therefore can effectively reject the interference of class area of skin color in the background, make the palm area of extraction more accurate, thereby improve the accuracy that checks finger tip, and in the process that detects finger tip, need not Fingers to directly over, also can correctly detect finger tip, enlarged range of application.
The embodiment of the invention also provides a kind of televisor at last, and this televisor comprises the finger tip pick-up unit of man-machine interaction.Wherein, the detailed construction of the finger tip pick-up unit of man-machine interaction no longer repeats at this referring to the associated description of above-described embodiment five to embodiment ten.
The above only is preferred embodiment of the present invention, not in order to limiting the present invention, all any modifications of doing within the spirit and principles in the present invention, is equal to and replaces and improvement etc., all should be included within protection scope of the present invention.

Claims (11)

1. the Fingertip Detection of a man-machine interaction is characterized in that, described method comprises the steps:
Obtain the area-of-interest of the video frame image of input, extract binaryzation movable information and the binaryzation Skin Color Information of described area-of-interest;
Divide palm area according to binaryzation movable information and binaryzation Skin Color Information, obtain the bianry image that comprises described palm area;
Obtain the centre coordinate of palm area in the described bianry image, and take described centre coordinate as the center of circle, default radius of circle length is that radius is justified, described radius of circle length is determined according to default initial length and default iteration step length;
Obtain the pixel value of pixel on the round path, and detect finger tip according to the pixel value that obtains, be specially:
Obtain one by one the pixel value of pixel on the round path;
It is 1 o'clock at the pixel value that obtains, whether the pixel value of judging the left and right sides neighbor of described pixel all is 0, the pixel value that at pixel value is the left and right sides neighbor of 1 pixel all is 0 o'clock, the subregion of judging described pixel place is finger tip, if not, the subregion of judging described pixel place is not finger tip;
Be 0 o'clock at the pixel value that obtains, continue to obtain the pixel value of the next pixel on the round path, until obtain the pixel value of all pixels on the round path;
Wherein, described is that the pixel value of the left and right sides neighbor of 1 pixel is 0 o'clock all at pixel value, and the subregion of judging described pixel place is specially as the step of finger tip:
The pixel value that at pixel value is the left and right sides neighbor of 1 pixel all is 0 o'clock, and centered by 1 pixel, default expansion multiple is expanded, and the zone after will expanding is divided into 4 zones take pixel value;
Detect the pixel value of pixel in 4 zones, if the pixel value of pixel all is 1 in one of them zone, and the pixel value of all the other regional pixels is 0, judges that then the central area in the zone after expanding is finger tip.
2. the method for claim 1 is characterized in that, the described area-of-interest that obtains the video frame image of input extracts the binaryzation movable information of described area-of-interest and the step of binaryzation Skin Color Information and is specially:
Determine the area-of-interest of video frame image;
The area-of-interest of video frame image is converted to gray level image, and adopts the three-frame difference method to detect the motion target area of gray level image, extract the binaryzation movable information according to described motion target area again;
According to default conversion formula the area-of-interest that obtains is converted to hue saturation value HSV color space from the RGB rgb color space, extracts the Skin Color Information of described HSV color space, extract the binaryzation Skin Color Information according to described Skin Color Information again.
3. the method for claim 1 is characterized in that, described according to binaryzation movable information and binaryzation Skin Color Information division palm area, the step of obtaining the bianry image that comprises described palm area is specially:
Divide palm area according to binaryzation movable information and binaryzation Skin Color Information;
Bianry image corresponding to palm area of dividing is divided into a plurality of subregions, and a plurality of subregions of dividing carried out morphology process.
4. method as claimed in claim 3 is characterized in that, described bianry image corresponding to palm area with division is divided into a plurality of subregions, and the step that a plurality of subregions execution morphology of dividing are processed is specially:
The bianry image that comprises palm area that obtains is divided into default a plurality of subregions, and a plurality of subregions of dividing are re-executed binary conversion treatment;
Adopt predefined nuclear to eliminate the soliton zone of palm area, perhaps fill the little gap of palm area.
5. method as claimed in claim 1 or 2, it is characterized in that, the described centre coordinate that obtains radius of circle length and palm area, and take described centre coordinate as the center of circle, the radius of circle length of obtaining is that radius is justified, and described radius of circle length is specially according to default initial length and the default definite step of iteration step length:
Determine radius of circle length by default initial length and iteration step length addition, or subtract each other definite radius of circle length by default initial length and iteration step length;
The pixel value of point by point scanning palm area is determined the centre coordinate of palm area according to the pixel value that obtains and pixel value position;
Take the centre coordinate of palm area as the center of circle, justify take the described radius of circle length determined as radius.
6. the method for claim 1 is characterized in that, described is that the pixel value of the left and right sides neighbor of 1 pixel is 0 o'clock all at pixel value, judges after the step of subregion as finger tip at described pixel place, further comprises the steps:
Whether judge the distance of the subregion be judged to be finger tip and the left and right sides adjacent area of described subregion all greater than the distance threshold of presetting, the left and right sides adjacent area of described subregion and the pixel value of the pixel between the described subregion are 0;
During all greater than default distance threshold, reject the described finger tip that has been judged to be finger tip in the distance of the left and right sides of the subregion that is judged to be finger tip and described subregion adjacent area.
7. the finger tip pick-up unit of a man-machine interaction is characterized in that, described device comprises:
The area-of-interest acquiring unit for the area-of-interest of the video frame image that obtains input, extracts binaryzation movable information and the binaryzation Skin Color Information of described area-of-interest;
The palm area acquiring unit is used for dividing palm area according to binaryzation movable information and binaryzation Skin Color Information, obtains the bianry image that comprises described palm area;
Circle path acquiring unit, be used for obtaining the centre coordinate of described bianry image palm area, and take described centre coordinate as the center of circle, default radius of circle length is that radius is justified, and described radius of circle length is determined according to default initial length and default iteration step length;
The finger tip detecting unit, for the pixel value that obtains pixel on the round path, and according to the pixel value detection finger tip that obtains;
Described finger tip detecting unit comprises:
The pixel value acquisition module is for the pixel value that obtains one by one pixel on the round path;
The first finger tip judge module, being used at the pixel value that obtains is 1 o'clock, whether the pixel value of judging the left and right sides neighbor of described pixel all is 0, the pixel value that at pixel value is the left and right sides neighbor of 1 pixel all is 0 o'clock, the subregion of judging described pixel place is finger tip, if not, the subregion of judging described pixel place is not finger tip;
Trigger module, being used at the pixel value that obtains is 0 o'clock, triggers described pixel value acquisition module and continues to obtain the pixel value of the next pixel on the round path, until obtain the pixel value of all pixels on the round path;
Wherein, described the first finger tip judge module specifically is used for:
The pixel value that at pixel value is the left and right sides neighbor of 1 pixel all is 0 o'clock, and centered by 1 pixel, default expansion multiple is expanded, and the zone after will expanding is divided into 4 zones take pixel value;
Detect the pixel value of pixel in 4 zones, if the pixel value of pixel all is 1 in one of them zone, and the pixel value of all the other regional pixels is 0, judges that then the central area in the zone after expanding is finger tip.
8. device as claimed in claim 7 is characterized in that, described palm area acquiring unit comprises:
Two-value palm area determination module is used for dividing palm area according to binaryzation movable information and binaryzation Skin Color Information;
The denoising module, bianry image corresponding to palm area that is used for dividing is divided into a plurality of subregions, and a plurality of subregions of dividing carried out morphology process.
9. device as claimed in claim 7 is characterized in that, described round path acquiring unit comprises:
The radius of circle determination module is used for determining radius of circle length by default initial length and iteration step length addition, or subtracts each other definite radius of circle length by default initial length and iteration step length;
Center of circle determination module is used for the pixel value of point by point scanning palm area, determines the centre coordinate of palm area according to the pixel value that obtains and pixel value position;
Circle path determination module is used for centre coordinate take palm area as the center of circle, justifies take the described radius of circle length determined as radius.
10. device as claimed in claim 7 is characterized in that, also comprises:
Whether the region distance judging unit be used for judges the distance of the subregion that has been judged to be finger tip and the left and right sides adjacent area of described subregion all greater than the distance threshold of presetting, and the left and right sides adjacent area of described subregion and the pixel value of the pixel between this subregion are 0;
The finger tip culling unit is used for during all greater than the distance threshold preset, rejecting the finger tip that this has been judged to be finger tip in the distance of the left and right sides of the subregion that is judged to be finger tip and this subregion adjacent area.
11. a televisor is characterized in that, described televisor comprises the finger tip pick-up unit of the described man-machine interaction of 7 to 10 each claims.
CN 201110388105 2011-11-29 2011-11-29 Human-computer interaction fingertip detection method, device and television Expired - Fee Related CN102521567B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN 201110388105 CN102521567B (en) 2011-11-29 2011-11-29 Human-computer interaction fingertip detection method, device and television

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN 201110388105 CN102521567B (en) 2011-11-29 2011-11-29 Human-computer interaction fingertip detection method, device and television

Publications (2)

Publication Number Publication Date
CN102521567A CN102521567A (en) 2012-06-27
CN102521567B true CN102521567B (en) 2013-10-23

Family

ID=46292478

Family Applications (1)

Application Number Title Priority Date Filing Date
CN 201110388105 Expired - Fee Related CN102521567B (en) 2011-11-29 2011-11-29 Human-computer interaction fingertip detection method, device and television

Country Status (1)

Country Link
CN (1) CN102521567B (en)

Families Citing this family (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105095838B (en) * 2014-05-20 2019-09-17 中国移动通信集团广东有限公司 A kind of object detection method and device
CN108230328B (en) 2016-12-22 2021-10-22 新沂阿凡达智能科技有限公司 Method and device for acquiring target object and robot
CN107341811B (en) * 2017-06-20 2020-11-13 上海数迹智能科技有限公司 Method for segmenting hand region by using MeanShift algorithm based on depth image
CN107678551B (en) * 2017-10-19 2021-12-28 京东方科技集团股份有限公司 Gesture recognition method and device and electronic equipment
CN108229391B (en) * 2018-01-02 2021-12-24 京东方科技集团股份有限公司 Gesture recognition device, server thereof, gesture recognition system and gesture recognition method
CN108921828B (en) * 2018-06-15 2022-04-22 湖南科技大学 Method for identifying insignificant weld joint in complex scene
CN111639562B (en) * 2020-05-15 2023-06-20 圣点世纪科技股份有限公司 Intelligent positioning method for palm region of interest
CN112232332B (en) * 2020-12-17 2021-04-13 四川圣点世纪科技有限公司 Non-contact palm detection method based on video sequence

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101694694A (en) * 2009-10-22 2010-04-14 上海交通大学 Finger identification method used in interactive demonstration system
CN101807114A (en) * 2010-04-02 2010-08-18 浙江大学 Natural interactive method based on three-dimensional gestures
CN102200834A (en) * 2011-05-26 2011-09-28 华南理工大学 television control-oriented finger-mouse interaction method

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
TWI398818B (en) * 2009-06-30 2013-06-11 Univ Nat Taiwan Science Tech Method and system for gesture recognition

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101694694A (en) * 2009-10-22 2010-04-14 上海交通大学 Finger identification method used in interactive demonstration system
CN101807114A (en) * 2010-04-02 2010-08-18 浙江大学 Natural interactive method based on three-dimensional gestures
CN102200834A (en) * 2011-05-26 2011-09-28 华南理工大学 television control-oriented finger-mouse interaction method

Also Published As

Publication number Publication date
CN102521567A (en) 2012-06-27

Similar Documents

Publication Publication Date Title
CN102521567B (en) Human-computer interaction fingertip detection method, device and television
US8989448B2 (en) Moving object detecting device, moving object detecting method, moving object detection program, moving object tracking device, moving object tracking method, and moving object tracking program
CN107272883B (en) Gesture preprocessing of video streams using tagged regions
US10489632B2 (en) Electronic device and detection method
CN109086747B (en) Gesture pre-processing of video streams using skin tone detection
US20170124719A1 (en) Method, device and computer-readable medium for region recognition
US20170140210A1 (en) Image processing apparatus and image processing method
CN105320248A (en) Mid-air gesture input method and device
EP3702956A1 (en) Gesture detection method, gesture processing device, and computer readable storage medium
US20130328773A1 (en) Camera-based information input method and terminal
CN105046254A (en) Character recognition method and apparatus
CN104966016B (en) Mobile terminal child user cooperatively judges and the method for limitation operating right
CN104318558A (en) Multi-information fusion based gesture segmentation method under complex scenarios
CN104809387A (en) Video image gesture recognition based non-contact unlocking method and device
CN108596944A (en) A kind of method, apparatus and terminal device of extraction moving target
CN104766354B (en) The method and mobile terminal that a kind of augmented reality is drawn
US20150035778A1 (en) Display control device, display control method, and computer program product
JP2012027745A (en) Information processing method and its device and program
CN114402369A (en) Human body posture recognition method and device, storage medium and electronic equipment
CN105389541A (en) Fingerprint image recognition method and device
CN104932683A (en) Game motion sensing control method based on vision information
Panwar Hand gesture based interface for aiding visually impaired
CN112329646A (en) Hand gesture motion direction identification method based on mass center coordinates of hand
CN109375833B (en) Touch instruction generation method and device
Hartanto et al. Real time hand gesture movements tracking and recognizing system

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

Granted publication date: 20131023

CF01 Termination of patent right due to non-payment of annual fee