CN103870071A - Touch source identification method and system - Google Patents
Touch source identification method and system Download PDFInfo
- Publication number
- CN103870071A CN103870071A CN201210541504.5A CN201210541504A CN103870071A CN 103870071 A CN103870071 A CN 103870071A CN 201210541504 A CN201210541504 A CN 201210541504A CN 103870071 A CN103870071 A CN 103870071A
- Authority
- CN
- China
- Prior art keywords
- pixel
- touch
- touch sources
- image
- sources
- 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.)
- Granted
Links
Images
Abstract
The invention discloses a touch source identification method. The method comprises acquiring a sensing image; according to the sensing image, removing noise and determining a touch source image; according to the touch source image and default touch source characteristics, determining the type of touch source. The invention also discloses a touch source identification system. According to the plan of the touch source identification method and system, a portable device with a capacitive touch screen can separate fingers from other common touch sources, such as cheeks, ears and hands, so that the identification accuracy is higher, and occurrences of error touch can be avoided effectively. Furthermore, the touch source identification method and system does not require using additional originals, such as proximity sensors and infrared light emitting diodes, so that the production cost and the difficulty of design are lower.
Description
Technical field
The present invention relates to mode identification technology, relate in particular to a kind of touch sources recognition methods and system.
Background technology
It is the main input mode of current smart mobile phone, panel computer by finger touch input.Touch-screen is the electric current induction that has utilized human body, user and touch screen surface form with a coupling capacitance, for high-frequency current, electric capacity is direct conductor, so finger siphons away a very little electric current from contact point, this electric current divides in the electrode four jiaos from touch-screen and flows out, and the electric current of these four electrodes of flowing through is directly proportional to the distance of four jiaos with pointing, controller, by the accurate Calculation to these four current ratios, draws the position of touch point.
But in communication process, other object maloperation touchscreen button such as cheek, ear cause mobile phone communication generation obstacle, the situations such as accidental interruption of even conversing often have generation, if portable equipment can be by finger and other common touch sources, as cheek, ear, palm etc. make a distinction, just can be by ignoring the operation of the touch sources beyond finger, to avoid false touch to occur.Even, the operating effect that intelligent and portable equipment can carry out the touch sources beyond finger is expanded the definition making new advances, and makes the mode of man-machine interaction more rich and varied.
At present, handheld device prevents that in communication process the non-finger touch sources such as user face from carrying out the main solution of maloperation and being: use a proximity transducer and infraluminescence pipe, judge the distance of user's head and mobile phone, if distance is less than 1～2cm, lock-screen.But, if adopt this by the cooperating of proximity transducer and infraluminescence pipe, control mobile phone screen locking and release by the distance on a bit, to prevent the mode of false touch in call, extra components and parts have not only increased appearance design difficulty and the production cost of mobile phone, and only sample on one point, if user's cheek and screen are also not parallel, probably fail to cause screen locking, and then false touch occurs, thereby can not avoid the generation of false touch completely.
In addition, for distinguishing of touch sources, application number is that 200780049219.9 patented claim also proposes a kind of settling mode: approach image by acquisition, cut apart and approach image to identify multiple blocks, determine each minor axis radius of these multiple blocks, for example, if the minor axis radius value of a block, on the first assign thresholds, is identified as large object (cheek) by this block, the operation of the large object control touch-surface equipment based on identified.But block is cut apart the accuracy that depends on watershed algorithm in this scheme, the self-defined empirical judgement depending in experiment of threshold value, thus these two links all exist the not factor of robust, make the recognition accuracy of the method be difficult to guarantee.
Summary of the invention
In view of this, fundamental purpose of the present invention is to provide a kind of touch sources recognition methods and system, and recognition accuracy is higher, can effectively avoid the generation of false touch, and design difficulty and production cost lower.
For achieving the above object, technical scheme of the present invention is achieved in that
A kind of touch sources recognition methods, comprising:
Obtain sensed image;
According to described sensed image, remove noise and determine touch sources image;
According to described touch sources image and default touch sources feature, determine the type of touch sources.
Described according to described sensed image, remove noise and determine that touch sources image is:
The pixel value of determining each pixel in sensed image is the distance of this some place touch sources apart from touch-screen;
The pixel that pixel value in sensed image is less than to default segmentation threshold is defined as foreground pixel, and the pixel adjacent with described foreground pixel determined the foreground pixel that is as the criterion, and other pixels are defined as background pixel;
Obtain the intermediate value of the pixel value of accurate foreground pixel and the pixel value of neighbor thereof, and the pixel value of described accurate foreground pixel is revised as to described intermediate value;
The accurate foreground pixel that pixel value is less than to segmentation threshold is defined as foreground pixel, and the accurate foreground pixel that pixel value is not less than segmentation threshold is defined as background pixel, so far, and all foreground pixel composition touch sources images.
Described according to described touch sources image and default touch sources feature, determine that the type of touch sources is:
Calculate seven Hu squares of described touch sources image, and described touch sources image is carried out to Skeleton processing, obtain target skeleton;
Calculate the mean curvature of described target skeleton;
To the mean curvature of described Hu square and described target skeleton, vote respectively according to default threshold value, touch sources is identified as to the highest classification of ballot value.
Described described touch sources image is carried out to Skeleton processing, obtains target skeleton and be:
A, the pixel value of each pixel in touch sources image is designated as to the inverse of this some place touch sources apart from the distance of touch-screen;
B, the pixel adjacent with background pixel and the non-background pixel that is positioned at touch-screen edge are defined as to contour pixel;
C, judge in touch sources image whether only have contour pixel, if so, current contour pixel composition target skeleton; Otherwise, execution step d;
D, traversal contour pixel, get minimum pixel value, deducts the value of described minimum pixel value as the new pixel value of this contour pixel using the pixel value of each contour pixel, and the contour pixel that new pixel value is zero is defined as background pixel, returns to afterwards step b.
Described according to described sensed image, before removing noise and determining touch sources image, the method also comprises:
According to the length at the area of touch sources contact screen and/or contact touch-screen edge, and default decision tree threshold value carries out decision tree classification, to determine candidate's touch sources or exclusive segment or whole non-object touch sources,
Wherein, carry out decision tree classification, while determining candidate's touch sources,
Described according to described touch sources image and default touch sources feature, determine that the type of touch sources is: according to described touch sources image and default touch sources feature corresponding to described candidate's touch sources, determine the type of touch sources.
A kind of touch sources recognition system, comprising: sensed image acquisition module, denoising and touch sources image determination module and touch sources type determination module; Wherein,
Described sensed image acquisition module, for obtaining sensed image;
Described denoising and touch sources image determination module, for according to described sensed image, remove noise and determine touch sources image;
Described touch sources type determination module, for according to described touch sources image and default touch sources feature, determines the type of touch sources.
Described denoising and touch sources image determination module specifically for:
The pixel value of determining each pixel in sensed image is the distance of this some place touch sources apart from touch-screen;
The pixel that pixel value in sensed image is less than to default segmentation threshold is defined as foreground pixel, and the pixel adjacent with described foreground pixel determined the foreground pixel that is as the criterion, and other pixels are defined as background pixel;
Obtain the intermediate value of the pixel value of accurate foreground pixel and the pixel value of neighbor thereof, and the pixel value of described accurate foreground pixel is revised as to described intermediate value;
The accurate foreground pixel that pixel value is less than to segmentation threshold is defined as foreground pixel, and the accurate foreground pixel that pixel value is not less than segmentation threshold is defined as background pixel, so far, and all foreground pixel composition touch sources images.
Described touch sources type determination module specifically for:
Calculate seven Hu squares of described touch sources image, and described touch sources image is carried out to Skeleton processing, obtain target skeleton;
Calculate the mean curvature of described target skeleton;
To the mean curvature of described Hu square and described target skeleton, vote respectively according to default threshold value, touch sources is identified as to the highest classification of ballot value.
Described touch sources type determination module is specifically for carrying out:
A, the pixel value of each pixel in touch sources image is designated as to the inverse of this some place touch sources apart from the distance of touch-screen;
B, the pixel adjacent with background pixel and the non-background pixel that is positioned at touch-screen edge are defined as to contour pixel;
C, judge in touch sources image whether only have contour pixel, if so, current contour pixel composition target skeleton; Otherwise, execution step d;
D, traversal contour pixel, get minimum pixel value, deducts the value of described minimum pixel value as the new pixel value of this contour pixel using the pixel value of each contour pixel, and the contour pixel that new pixel value is zero is defined as background pixel, returns to afterwards step b.
This system also comprises decision tree classification module,
Described decision tree classification module, for obtaining after sensed image at sensed image acquisition module, according to the length at the area of touch sources contact screen and/or contact touch-screen edge, and default decision tree threshold value is carried out decision tree classification, to determine candidate's touch sources or exclusive segment or whole non-object touch sources
Wherein, carry out decision tree classification, while determining candidate's touch sources,
Described according to described touch sources image and default touch sources feature, determine that the type of touch sources is: according to described touch sources image and default touch sources feature corresponding to described candidate's touch sources, determine the type of touch sources.
Touch sources recognition methods of the present invention and system, obtain sensed image; According to described sensed image, remove noise and determine touch sources image; According to described touch sources image and default touch sources feature, determine the type of touch sources.According to scheme of the present invention, can make will point and other common touch sources with the portable equipment of capacitance touch screen, as cheek, ear, palm etc. distinguish, thereby recognition accuracy is higher, can effectively avoid the generation of false touch, and the present invention does not need to use the extra original paper such as proximity transducer, infraluminescence pipe, thereby design difficulty and production cost are lower.
Accompanying drawing explanation
Fig. 1 is a kind of touch sources recognition methods of one embodiment of the invention schematic flow sheet;
Fig. 2 is the sensed image schematic diagram that one embodiment of the invention is obtained;
Fig. 3 is one embodiment of the invention is removed noise definite touch sources image detailed process schematic diagram according to sensed image;
Fig. 4 is foreground pixel and the background pixel schematic diagram obtaining in one embodiment of the invention;
Fig. 5 carries out pixel schematic diagram after treatment based on Fig. 4 in one embodiment of the invention;
Fig. 6 carries out pixel schematic diagram after treatment based on Fig. 5 in one embodiment of the invention;
Fig. 7 be in one embodiment of the invention according to touch sources image and default touch sources feature, determine the detailed process schematic diagram of the type of touch sources;
Fig. 8 is the overall flow schematic diagram of one embodiment of the invention;
Fig. 9 is common sensed image schematic diagram;
Figure 10 carries out Skeleton processing to touch sources image in one embodiment of the invention, obtains the detailed process schematic diagram of target skeleton;
Figure 11 is the sensed image schematic diagram in one embodiment of the invention;
Figure 12 is through step a and step b pixel schematic diagram after treatment in one embodiment of the invention;
Figure 13 carries out based on Figure 12 the pixel situation schematic diagram obtaining after an iterative processing in one embodiment of the invention;
Figure 14 carries out repeatedly iterative processing based on Figure 12 in one embodiment of the invention, until the pixel situation schematic diagram obtaining while only having contour pixel in touch sources image;
Figure 15 is based on Figure 12, the pixel situation schematic diagram that adopts traditional Skeleton method to obtain;
Figure 16 is a kind of touch sources recognition methods of another embodiment of the present invention schematic flow sheet;
Figure 17 is the decision tree schematic diagram that one embodiment of the invention adopts;
Figure 18 is a kind of touch sources recognition system of one embodiment of the invention structural representation;
Figure 19 is a kind of touch sources recognition system of another embodiment of the present invention structural representation.
Embodiment
Basic thought of the present invention is: obtain sensed image; According to described sensed image, remove noise and determine touch sources image; According to described touch sources image and default touch sources feature, determine the type of touch sources.
The embodiment of the present invention has proposed a kind of touch sources recognition methods, and as shown in Figure 1, the method comprises:
Step 101: obtain sensed image;
Here, generally scan by the two-dimensional array of elements to touch-screen, obtain the capacitance variations value of multiple column electrodes and row electrode crossings, obtain sensed image in conjunction with the coordinate figure of multiple column electrodes and row electrode crossings.For example, the sensed image of obtaining as shown in Figure 2.
Step 102: according to described sensed image, remove noise and determine touch sources image;
For after identification to target, need to be to extracting touch sources part from sensed image in this step, and the induced noise of touch-screen is removed.Concrete, remove noise and determine touch sources image according to the pixel value of each pixel in sensed image (this some place touch sources is apart from the distance of touch-screen).
Step 103: according to described touch sources image and default touch sources feature, determine the type of touch sources.
Optionally, as shown in Figure 3, according to described sensed image, remove noise and determine that touch sources image comprises described in step 102:
Step 1021: the pixel value of determining each pixel in sensed image is the distance of this some place touch sources apart from touch-screen;
Step 1022: the pixel that pixel value in sensed image is less than to default segmentation threshold is defined as foreground pixel, other pixels are defined as background pixel;
Preferably, segmentation threshold is 5 millimeters, adopt and can guarantee that in this way important foreground pixel value keeps original value constant, aspect the prospect of removal noise, accomplish meticulousr, with respect to traditional morphological method and the more satisfied requirement on touch-screen of filtering operation.For example, as shown in Figure 4, the numerical value in figure is pixel value corresponding to each pixel for the foreground pixel that this step is obtained and background pixel.
Step 1023: the pixel adjacent with described foreground pixel determined to the foreground pixel that is as the criterion, obtain the intermediate value of the pixel value of accurate foreground pixel and the pixel value of neighbor thereof, and the pixel value of described accurate foreground pixel is revised as to described intermediate value;
For obtaining of intermediate value, for instance, if the pixel value of accurate foreground pixel is 5, the pixel value of the neighbor of this accurate foreground pixel is respectively 2,2,7,8, so, it is 2,2,5,7,8 that described five pixel values are sequentially sorted by size, and the intermediate value 5 of this sequence is the intermediate value of described five pixel values.Based on Fig. 4, carry out this and walk pixel schematic diagram after treatment as shown in Figure 5.
Step 1024: the accurate foreground pixel that pixel value is less than to segmentation threshold is defined as foreground pixel, the accurate foreground pixel that pixel value is not less than segmentation threshold is defined as background pixel, so far, all foreground pixel composition touch sources images.
Based on Fig. 5, carry out this and walk pixel schematic diagram after treatment as shown in Figure 6.
Optionally, as shown in Figure 7, according to described touch sources image and default touch sources feature, determine that the type of touch sources comprises described in step 103:
Step 1031: calculate seven Hu squares of described touch sources image, and described touch sources image is carried out to Skeleton processing, obtain target skeleton;
Step 1032: the mean curvature of calculating described target skeleton;
Step 1033: to the mean curvature of described Hu square and described target skeleton, vote respectively according to default threshold value, touch sources is identified as to the highest classification of ballot value.
It should be noted that, the present invention has sorter training and two links of real-time grading, and as shown in Figure 8, wherein, sorter training link is link under line to overall flow, and training once.Sensed image (the sensed image as shown in Figure 9 that this link is used specific finger, ear, cheek to produce, wherein first row is finger sensed image, second row is cheek sensed image, the 3rd row is ear sensed image), Hu square and curvature feature when calculating standard finger, ear, cheek contact touch-screen, object is to obtain the one group of grouped data (i.e. default threshold value) that can distinguish common touch sources; Link is link on line in real time, and object is in the time that user uses touch-screen, according to sensed image, uses existing sorter, and finger touch is distinguished over to other object contact touch-screens.This link calculated amount is less, can reach Real time Efficiency.
Optionally, as shown in figure 10, described touch sources image is carried out to Skeleton processing described in step 1031, obtains target skeleton and comprise:
Step a: the pixel value of each pixel in touch sources image is designated as to the inverse of this some place touch sources apart from the distance of touch-screen;
Step b: the pixel adjacent with background pixel and the non-background pixel that is positioned at touch-screen edge are defined as to contour pixel;
For example, the sensed image shown in Figure 11 through step a and step b pixel schematic diagram after treatment as shown in figure 12, wherein because 0 inverse does not exist, so be labeled as m.
Step c: judge in touch sources image whether only have contour pixel, if so, current contour pixel composition target skeleton; Otherwise, execution step d;
For the dot structure shown in Figure 12, due in touch sources image except contour pixel, also comprise foreground pixel, therefore, execution step d.
Steps d: traversal contour pixel, get minimum pixel value, deduct the value of described minimum pixel value using the pixel value of each contour pixel as the new pixel value of this contour pixel, the contour pixel that new pixel value is zero is defined as background pixel, returns to afterwards step b.
Based on Figure 12, carry out the pixel situation obtained after an iterative processing as shown in figure 13, carry out repeatedly iterative processing, until the pixel situation of obtaining while only having contour pixel in touch sources image is as shown in figure 14, wherein, contour pixel composition target skeleton.
It should be noted that, the present invention be adopt improved image Skeleton be according to respective pixel value on object shapes and objective contour to target yojan, make result can embody the distance relation of object and touch-screen.Traditional Skeleton is to operate by continuous corrosion, is one group of thin bone by the object reduction in bianry image, thereby extracts target skeleton.Thin bone still retains the important information of primary object shape, based on Figure 12, adopts pixel situation schematic diagram that traditional Skeleton method obtains as shown in figure 15, can find out, the method can not meet requirement of the present invention.
In the present invention, touch sources can be finger, cheek, ear, palm, elbow etc.
Optionally, as shown in figure 16, according to described sensed image, before removing noise and determining touch sources image, the method also comprises described in step 102:
Step 102 ': according to the length at the area of touch sources contact screen and/or contact touch-screen edge, and default decision tree threshold value is carried out decision tree classification, to determine candidate's touch sources or exclusive segment or whole non-object touch sources, determine and select touch sources or may, for after object touch sources, perform step 102 at touch sources.
It should be noted that, carry out decision tree classification, while determining candidate's touch sources, described according to described touch sources image and default touch sources feature, the type of determining touch sources is: according to described touch sources image and default touch sources feature corresponding to described candidate's touch sources, determine the type of touch sources.
Optionally, the decision tree that this step adopts can be as shown in figure 17, for getting rid of non-object touch sources.Here, the meaning of decision tree classification is, 1) can identify fast the touch behavior in some non-finger touch sources, get rid of; 2) during due to contact touch-screens such as ear, palm, cheek, sensed image may can only be obtained their topography, so just brings larger uncertainty to follow-up recognition feature.Decision tree extracts the small size of touch sources in touch screen zone and touches behavior, is convenient to the identification of subsequent step.
Based on such scheme, when touch sources is during near the capacitance touch screen of portable equipment, the signal of the two-dimensional array output of touchscreen senses element is considered as to sensed image.By the signature analysis to sensed image, portable equipment will be pointed and other common touch sources, as cheek, ear, palm etc. distinguish, can reach the effect that prevents non-finger false touch.And, in real-time grading link, only have simple image to process and morphology calculating, calculated amount is very little, for internal memory and the computing power of current smart mobile phone, can reach real-time performance completely.
It should be noted that, the present invention can be widely used in, on the intelligent and portable equipment with capacitance touch screen, as intelligent and portable equipment, comprising mobile phone, panel computer etc.Can realize by the present invention: distinguish when touch sources is legal touch sources, allow to touch operation, otherwise will operate invalidly, avoided thus false touch and the maloperation in routine use; In the time distinguishing that touch sources is illegal touch sources, lock-screen is also shown as minimum brightness, saves smart machine electric weight; In addition, the present invention to following smart machine expand legal touch sources, the definition of enriching man-machine interaction mode is extremely helpful.
The embodiment of the present invention has also correspondingly proposed a kind of touch sources recognition system, and as shown in figure 18, this system comprises: sensed image acquisition module, denoising and touch sources image determination module and touch sources type determination module; Wherein,
Described sensed image acquisition module, for obtaining sensed image;
Described denoising and touch sources image determination module, for according to described sensed image, remove noise and determine touch sources image;
Described touch sources type determination module, for according to described touch sources image and default touch sources feature, determines the type of touch sources.
Optionally, described denoising and touch sources image determination module specifically for:
The pixel value of determining each pixel in sensed image is the distance of this some place touch sources apart from touch-screen;
The pixel that pixel value in sensed image is less than to default segmentation threshold is defined as foreground pixel, and the pixel adjacent with described foreground pixel determined the foreground pixel that is as the criterion, and other pixels are defined as background pixel;
Obtain the intermediate value of the pixel value of accurate foreground pixel and the pixel value of neighbor thereof, and the pixel value of described accurate foreground pixel is revised as to described intermediate value;
The accurate foreground pixel that pixel value is less than to segmentation threshold is defined as foreground pixel, and the accurate foreground pixel that pixel value is not less than segmentation threshold is defined as background pixel, so far, and all foreground pixel composition touch sources images.
Optionally, described touch sources type determination module specifically for:
Calculate seven Hu squares of described touch sources image, and described touch sources image is carried out to Skeleton processing, obtain target skeleton;
Calculate the mean curvature of described target skeleton;
To the mean curvature of described Hu square and described target skeleton, vote respectively according to default threshold value, touch sources is identified as to the highest classification of ballot value.
Optionally, described touch sources type determination module is specifically for carrying out:
A, the pixel value of each pixel in touch sources image is designated as to the inverse of this some place touch sources apart from the distance of touch-screen;
B, the pixel adjacent with background pixel and the non-background pixel that is positioned at touch-screen edge are defined as to contour pixel;
C, judge in touch sources image whether only have contour pixel, if so, current contour pixel composition target skeleton; Otherwise, execution step d;
D, traversal contour pixel, get minimum pixel value, deducts the value of described minimum pixel value as the new pixel value of this contour pixel using the pixel value of each contour pixel, and the contour pixel that new pixel value is zero is defined as background pixel, returns to afterwards step b.
Optionally, as shown in figure 19, this system also comprises decision tree classification module,
Described decision tree classification module, for obtaining after sensed image at sensed image acquisition module, according to the length at the area of touch sources contact screen and/or contact touch-screen edge, and default decision tree threshold value is carried out decision tree classification, to determine candidate's touch sources or exclusive segment or whole non-object touch sources
Wherein, carry out decision tree classification, while determining candidate's touch sources,
Described according to described touch sources image and default touch sources feature, determine that the type of touch sources is: according to described touch sources image and default touch sources feature corresponding to described candidate's touch sources, determine the type of touch sources.
In sum, the present invention possesses following technological merit:
1) by decision tree classification, identify fast the touch behavior in some non-finger touch sources, get rid of at once; 2) during due to contact touch-screens such as ear, palm, cheek, sensed image may can only be obtained their topography, so just brings larger uncertainty to follow-up recognition feature.Decision tree extracts the small size of touch sources in touch screen zone and touches behavior, is convenient to the identification of subsequent step;
2), by improved morphology computing method, remove induced noise and extract touch sources.The method can be extracted touch sources part from sensed image, and the induced noise of touch-screen is removed, the identification to target after being convenient to;
3) improved image Skeleton, to target yojan, makes result can embody the distance relation of object and touch-screen according to respective pixel value on target shape and objective contour.And in conjunction with utilizing Hu square and touch sources skeleton curvature feature fast finger and other touch sources to be distinguished.
The above, be only preferred embodiment of the present invention, is not intended to limit protection scope of the present invention.
Claims (10)
1. a touch sources recognition methods, is characterized in that, the method comprises:
Obtain sensed image;
According to described sensed image, remove noise and determine touch sources image;
According to described touch sources image and default touch sources feature, determine the type of touch sources.
2. method according to claim 1, is characterized in that, described according to described sensed image, removes noise and determines that touch sources image is:
The pixel value of determining each pixel in sensed image is the distance of this some place touch sources apart from touch-screen;
The pixel that pixel value in sensed image is less than to default segmentation threshold is defined as foreground pixel, and the pixel adjacent with described foreground pixel determined the foreground pixel that is as the criterion, and other pixels are defined as background pixel;
Obtain the intermediate value of the pixel value of accurate foreground pixel and the pixel value of neighbor thereof, and the pixel value of described accurate foreground pixel is revised as to described intermediate value;
The accurate foreground pixel that pixel value is less than to segmentation threshold is defined as foreground pixel, and the accurate foreground pixel that pixel value is not less than segmentation threshold is defined as background pixel, so far, and all foreground pixel composition touch sources images.
3. method according to claim 1, is characterized in that, described according to described touch sources image and default touch sources feature, determines that the type of touch sources is:
Calculate seven Hu squares of described touch sources image, and described touch sources image is carried out to Skeleton processing, obtain target skeleton;
Calculate the mean curvature of described target skeleton;
To the mean curvature of described Hu square and described target skeleton, vote respectively according to default threshold value, touch sources is identified as to the highest classification of ballot value.
4. method according to claim 3, is characterized in that, described described touch sources image is carried out to Skeleton processing, obtains target skeleton and is:
A, the pixel value of each pixel in touch sources image is designated as to the inverse of this some place touch sources apart from the distance of touch-screen;
B, the pixel adjacent with background pixel and the non-background pixel that is positioned at touch-screen edge are defined as to contour pixel;
C, judge in touch sources image whether only have contour pixel, if so, current contour pixel composition target skeleton; Otherwise, execution step d;
D, traversal contour pixel, get minimum pixel value, deducts the value of described minimum pixel value as the new pixel value of this contour pixel using the pixel value of each contour pixel, and the contour pixel that new pixel value is zero is defined as background pixel, returns to afterwards step b.
5. according to the method described in claim 1 to 4 any one, it is characterized in that, described according to described sensed image, before removing noise and determining touch sources image, the method also comprises:
According to the length at the area of touch sources contact screen and/or contact touch-screen edge, and default decision tree threshold value carries out decision tree classification, to determine candidate's touch sources or exclusive segment or whole non-object touch sources,
Wherein, carry out decision tree classification, while determining candidate's touch sources,
Described according to described touch sources image and default touch sources feature, determine that the type of touch sources is: according to described touch sources image and default touch sources feature corresponding to described candidate's touch sources, determine the type of touch sources.
6. a touch sources recognition system, is characterized in that, this system comprises: sensed image acquisition module, denoising and touch sources image determination module and touch sources type determination module; Wherein,
Described sensed image acquisition module, for obtaining sensed image;
Described denoising and touch sources image determination module, for according to described sensed image, remove noise and determine touch sources image;
Described touch sources type determination module, for according to described touch sources image and default touch sources feature, determines the type of touch sources.
7. system according to claim 6, is characterized in that, described denoising and touch sources image determination module specifically for:
The pixel value of determining each pixel in sensed image is the distance of this some place touch sources apart from touch-screen;
The pixel that pixel value in sensed image is less than to default segmentation threshold is defined as foreground pixel, and the pixel adjacent with described foreground pixel determined the foreground pixel that is as the criterion, and other pixels are defined as background pixel;
Obtain the intermediate value of the pixel value of accurate foreground pixel and the pixel value of neighbor thereof, and the pixel value of described accurate foreground pixel is revised as to described intermediate value;
The accurate foreground pixel that pixel value is less than to segmentation threshold is defined as foreground pixel, and the accurate foreground pixel that pixel value is not less than segmentation threshold is defined as background pixel, so far, and all foreground pixel composition touch sources images.
8. system according to claim 6, is characterized in that, described touch sources type determination module specifically for:
Calculate seven Hu squares of described touch sources image, and described touch sources image is carried out to Skeleton processing, obtain target skeleton;
Calculate the mean curvature of described target skeleton;
To the mean curvature of described Hu square and described target skeleton, vote respectively according to default threshold value, touch sources is identified as to the highest classification of ballot value.
9. system according to claim 8, is characterized in that, described touch sources type determination module is specifically for carrying out:
A, the pixel value of each pixel in touch sources image is designated as to the inverse of this some place touch sources apart from the distance of touch-screen;
B, the pixel adjacent with background pixel and the non-background pixel that is positioned at touch-screen edge are defined as to contour pixel;
C, judge in touch sources image whether only have contour pixel, if so, current contour pixel composition target skeleton; Otherwise, execution step d;
D, traversal contour pixel, get minimum pixel value, deducts the value of described minimum pixel value as the new pixel value of this contour pixel using the pixel value of each contour pixel, and the contour pixel that new pixel value is zero is defined as background pixel, returns to afterwards step b.
10. according to the system described in claim 6 to 9 any one, it is characterized in that, this system also comprises decision tree classification module,
Described decision tree classification module, for obtaining after sensed image at sensed image acquisition module, according to the length at the area of touch sources contact screen and/or contact touch-screen edge, and default decision tree threshold value is carried out decision tree classification, to determine candidate's touch sources or exclusive segment or whole non-object touch sources
Wherein, carry out decision tree classification, while determining candidate's touch sources,
Described according to described touch sources image and default touch sources feature, determine that the type of touch sources is: according to described touch sources image and default touch sources feature corresponding to described candidate's touch sources, determine the type of touch sources.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210541504.5A CN103870071B (en) | 2012-12-13 | 2012-12-13 | One kind touches source discrimination and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210541504.5A CN103870071B (en) | 2012-12-13 | 2012-12-13 | One kind touches source discrimination and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103870071A true CN103870071A (en) | 2014-06-18 |
| CN103870071B CN103870071B (en) | 2017-09-08 |
Family
ID=50908672
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201210541504.5A Active CN103870071B (en) | 2012-12-13 | 2012-12-13 | One kind touches source discrimination and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103870071B (en) |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104407793A (en) * | 2014-11-26 | 2015-03-11 | 深圳市华星光电技术有限公司 | Method and equipment for processing touch signal |
| CN105245716A (en) * | 2015-10-27 | 2016-01-13 | 青岛海信移动通信技术股份有限公司 | Method and mobile terminal for preventing from accidentally dialling |
| CN105389541A (en) * | 2015-10-19 | 2016-03-09 | 广东欧珀移动通信有限公司 | Fingerprint image recognition method and device |
| CN106059556A (en) * | 2015-04-17 | 2016-10-26 | 福特全球技术公司 | Proximity switch assemby with signal drift rejection and method |
| CN107592419A (en) * | 2015-10-19 | 2018-01-16 | 广东欧珀移动通信有限公司 | Display screen awakening method, device and mobile terminal based on fingerprint sensor |
| CN107808089A (en) * | 2017-11-21 | 2018-03-16 | 青岛海信移动通信技术股份有限公司 | A kind of fingerprint identification method, device and storage medium |
| CN109032432A (en) * | 2018-07-17 | 2018-12-18 | 深圳市天英联合教育股份有限公司 | A kind of method, apparatus and terminal device of lettering pen category identification |
| CN110109563A (en) * | 2018-02-01 | 2019-08-09 | 奇手公司 | A kind of method and system of the contact condition of determining object relative to touch sensitive surface |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20050226505A1 (en) * | 2004-03-31 | 2005-10-13 | Wilson Andrew D | Determining connectedness and offset of 3D objects relative to an interactive surface |
| CN1879077A (en) * | 2004-09-07 | 2006-12-13 | 索尼公司 | Information processing unit and method, and program |
| CN101464948A (en) * | 2009-01-14 | 2009-06-24 | 北京航空航天大学 | Object identification method for affine constant moment based on key point |
| CN101583924A (en) * | 2007-01-03 | 2009-11-18 | 苹果公司 | Multi-touch input discrimination |
-
2012
- 2012-12-13 CN CN201210541504.5A patent/CN103870071B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20050226505A1 (en) * | 2004-03-31 | 2005-10-13 | Wilson Andrew D | Determining connectedness and offset of 3D objects relative to an interactive surface |
| CN1879077A (en) * | 2004-09-07 | 2006-12-13 | 索尼公司 | Information processing unit and method, and program |
| CN101583924A (en) * | 2007-01-03 | 2009-11-18 | 苹果公司 | Multi-touch input discrimination |
| CN101464948A (en) * | 2009-01-14 | 2009-06-24 | 北京航空航天大学 | Object identification method for affine constant moment based on key point |
Non-Patent Citations (1)
| Title |
|---|
| 薛召军等: "基于多参数模型结合的步态特征提取新方法", 《中国生物医学工程学报》 * |
Cited By (17)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104407793B (en) * | 2014-11-26 | 2018-03-13 | 深圳市华星光电技术有限公司 | Touch signal processing method and equipment |
| CN104407793A (en) * | 2014-11-26 | 2015-03-11 | 深圳市华星光电技术有限公司 | Method and equipment for processing touch signal |
| WO2016082251A1 (en) * | 2014-11-26 | 2016-06-02 | 深圳市华星光电技术有限公司 | Touch signal processing method and device |
| CN106059556B (en) * | 2015-04-17 | 2022-01-11 | 福特全球技术公司 | Proximity switch assembly with signal drift suppression and method |
| CN106059556A (en) * | 2015-04-17 | 2016-10-26 | 福特全球技术公司 | Proximity switch assemby with signal drift rejection and method |
| CN107592419A (en) * | 2015-10-19 | 2018-01-16 | 广东欧珀移动通信有限公司 | Display screen awakening method, device and mobile terminal based on fingerprint sensor |
| US10755076B2 (en) | 2015-10-19 | 2020-08-25 | Guangdong Oppo Mobile Telecommunications Corp., Ltd. | Fingerprint image recognition method, apparatus and terminal |
| WO2017067270A1 (en) * | 2015-10-19 | 2017-04-27 | 广东欧珀移动通信有限公司 | Fingerprint image recognition method, apparatus, and terminal |
| CN105389541A (en) * | 2015-10-19 | 2016-03-09 | 广东欧珀移动通信有限公司 | Fingerprint image recognition method and device |
| CN105389541B (en) * | 2015-10-19 | 2018-05-01 | 广东欧珀移动通信有限公司 | The recognition methods of fingerprint image and device |
| CN108427932A (en) * | 2015-10-19 | 2018-08-21 | 广东欧珀移动通信有限公司 | The recognition methods of fingerprint image and device |
| CN108427932B (en) * | 2015-10-19 | 2021-07-02 | Oppo广东移动通信有限公司 | Fingerprint image identification method and device |
| US10572714B2 (en) | 2015-10-19 | 2020-02-25 | Guangdong Oppo Mobile Telecommunications Corp., Ltd. | Fingerprint image recognition method, apparatus and terminal |
| CN105245716A (en) * | 2015-10-27 | 2016-01-13 | 青岛海信移动通信技术股份有限公司 | Method and mobile terminal for preventing from accidentally dialling |
| CN107808089A (en) * | 2017-11-21 | 2018-03-16 | 青岛海信移动通信技术股份有限公司 | A kind of fingerprint identification method, device and storage medium |
| CN110109563A (en) * | 2018-02-01 | 2019-08-09 | 奇手公司 | A kind of method and system of the contact condition of determining object relative to touch sensitive surface |
| CN109032432A (en) * | 2018-07-17 | 2018-12-18 | 深圳市天英联合教育股份有限公司 | A kind of method, apparatus and terminal device of lettering pen category identification |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103870071B (en) | 2017-09-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103870071A (en) | Touch source identification method and system | |
| CN103294996B (en) | A kind of 3D gesture identification method | |
| RU2711029C2 (en) | Touch classification | |
| CN102063618B (en) | Dynamic gesture identification method in interactive system | |
| CN103065134B (en) | A kind of fingerprint identification device and method with information | |
| CN102831404B (en) | Gesture detecting method and system | |
| CN103971102A (en) | Static gesture recognition method based on finger contour and decision-making trees | |
| TWI435276B (en) | A method and apparatus for recognition of handwritten symbols | |
| CN102402289B (en) | Mouse recognition method for gesture based on machine vision | |
| Jinda-Apiraksa et al. | A simple shape-based approach to hand gesture recognition | |
| RU2013154102A (en) | FINGER RECOGNITION AND TRACKING SYSTEM | |
| TWI528271B (en) | Method, apparatus and computer program product for polygon gesture detection and interaction | |
| WO2013051681A1 (en) | Finger shape estimation device, finger shape estimation method, and finger shape estimation program | |
| CN111414837A (en) | Gesture recognition method and device, computer equipment and storage medium | |
| Vivek Veeriah et al. | Robust hand gesture recognition algorithm for simple mouse control | |
| CN109189219A (en) | The implementation method of contactless virtual mouse based on gesture identification | |
| CN107037951B (en) | Automatic operation mode identification method and terminal | |
| CN102194097A (en) | Multifunctional method for identifying hand gestures | |
| KR101967858B1 (en) | Apparatus and method for separating objects based on 3D depth image | |
| CN111492407A (en) | System and method for drawing beautification | |
| CN106650554A (en) | Static hand gesture identification method | |
| KR101167784B1 (en) | A method for recognizing pointers and a method for recognizing control commands, based on finger motions on the back of the portable information terminal | |
| CN111488897B (en) | Method and device for detecting and identifying touch object | |
| Ying et al. | Fingertip detection and tracking using 2D and 3D information | |
| CN111639573B (en) | ORB algorithm-based gesture recognition method, storage medium and electronic equipment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |