CN103514437B - A kind of three-dimension gesture identifying device and three-dimensional gesture recognition method - Google Patents
A kind of three-dimension gesture identifying device and three-dimensional gesture recognition method Download PDFInfo
- Publication number
- CN103514437B CN103514437B CN201210565201.7A CN201210565201A CN103514437B CN 103514437 B CN103514437 B CN 103514437B CN 201210565201 A CN201210565201 A CN 201210565201A CN 103514437 B CN103514437 B CN 103514437B
- Authority
- CN
- China
- Prior art keywords
- gesture
- image
- emitting diode
- dimensional
- infrarede emitting
- 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.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 28
- 230000008569 process Effects 0.000 claims abstract description 9
- 230000004807 localization Effects 0.000 claims description 22
- 210000004247 hand Anatomy 0.000 claims description 11
- 238000006243 chemical reaction Methods 0.000 claims description 10
- 230000000007 visual effect Effects 0.000 claims description 8
- 230000003628 erosive effect Effects 0.000 claims description 6
- 230000001154 acute effect Effects 0.000 claims description 3
- 238000011084 recovery Methods 0.000 claims description 3
- 230000011218 segmentation Effects 0.000 claims description 3
- 238000005286 illumination Methods 0.000 abstract description 2
- 238000010586 diagram Methods 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 4
- 230000003993 interaction Effects 0.000 description 3
- 230000004913 activation Effects 0.000 description 2
- 230000007423 decrease Effects 0.000 description 2
- 230000007547 defect Effects 0.000 description 2
- 230000007812 deficiency Effects 0.000 description 2
- 238000010276 construction Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
Landscapes
- Length Measuring Devices By Optical Means (AREA)
- User Interface Of Digital Computer (AREA)
Abstract
A kind of three-dimension gesture identifying device and three-dimensional gesture recognition method, identifying device include two infrared cameras, image acquisition units, infrarede emitting diode, infrarede emitting diode driver element, arithmetic element and gesture identification unit；Two infrared photographies are head-up arranged and are electrically connected with image acquisition units, and infrarede emitting diode driver element drives infrarede emitting diode to light, and arithmetic element is electrically connected with image acquisition units, infrarede emitting diode driver element, gesture identification unit；Infrarede emitting diode is luminous to provide bias light, and image acquisition units gather images of gestures Data Concurrent by infrared camera and are sent to arithmetic element and carry out binaryzation, edge extracting, two-dimentional finger locating, are sent to gesture identification unit after three-positional fix process and realize three-dimension gesture identification.The present invention adopts infrared illumination and infrared camera, solves visible ray interference, and photographic head intake image upward reduces the interference of background, further increases the reliability of gesture identification.
Description
Technical field
The present invention relates to Gesture Recognition, more particularly to a kind of three-dimension gesture identifying device three-dimensional gesture recognition method.
Background technology
At present, gesture interaction technology has been widely used for field of human-computer interaction, and prior art gesture interaction technology is main
Recognize including the two-dimentional gesture identification based on single camera and the photographic head three-dimension gesture based on infrared structure light, the latter such as Microsoft
The product kinect of company.Depth information is operated without based on the two-dimentional gesture identification of single camera, can only identification letter
Single two-dimentional gesture information；Three-dimension gesture identification cost of implementation based on infrared structure light is higher, and performance accuracy has certain defect.
Therefore, prior art could be improved and develop.
The content of the invention
In view of in place of above-mentioned the deficiencies in the prior art, the present invention proposes a kind of base to solve prior art defect and deficiency
In the three-dimension gesture identification equipment and gesture identification method of common infrared camera, the interference of environment can be eliminated, improve gesture and know
Other reliability, calculating process are simple, and cost of implementation is low.
The technical proposal for solving the technical problem of the invention is as follows：
A kind of three-dimension gesture identifying device, for being identified to gesture, including two infrared cameras, image acquisition list
Unit, infrarede emitting diode, infrarede emitting diode driver element, arithmetic element and gesture identification unit；It is described two infrared
The camera lens of photographic head is arranged upward, and described image collecting unit is electrically connected with two infrared cameras, two pole of the infraluminescence
Pipe driver element is electrically connected with infrarede emitting diode to drive infrarede emitting diode to light, and the arithmetic element is adopted with image
Collection unit, infrarede emitting diode driver element, gesture identification unit electrical connection；Two pole of arithmetic element control infraluminescence
The images of gestures that pipe lights as the intake of two infrared cameras provides bias light, and described image collecting unit passes through two and infrared takes the photograph
As head gathers images of gestures data, and images of gestures data is activation is processed to arithmetic element, arithmetic element obtains gesture
Gesture identification unit is sent to after view data and realizes that three-dimension gesture is recognized.
Horizontal range between described two infrared cameras is 6cm.
The visual angle angle of described two infrared cameras is 60 °.
The infrared diode is at least one.
Present invention also offers a kind of three-dimensional gesture recognition method, is realized based on above-mentioned three-dimension gesture identifying device, its
It is characterised by, comprises the steps：
Left and right two infrared cameras are horizontally disposed with, and cam lens are upward；
Described image collecting unit obtains visual angle above two infrared cameras respectively by left and right two infrared cameras
In the range of first gesture image, second gesture image sent to arithmetic element；
First gesture image, second gesture image are carried out image binaryzation process by arithmetic element respectively；
First gesture image after binary conversion treatment, second gesture image are carried out Boundary extracting algorithm by arithmetic element respectively
Obtain edge image, the edge image of second gesture image of first gesture image；
Arithmetic element realizes finger two dimension according to the edge image of first gesture image, the edge image of second gesture image
Coordinate setting；
Arithmetic element is by the edge graph of the two-dimensional localization coordinate and second gesture image of the edge image of first gesture image
The two-dimensional localization Coordinate Conversion of picture is entered into three-dimensional localization coordinate with the gesture model information being pre-stored within gesture identification unit
Row match cognization.
Used as further improved scheme, gesture operation scope is the common factor at the visual angle of two infrared cameras.
The Binarization methods are realized using common Threshold segmentation.
The Boundary extracting algorithm first removes spuious point using the erosion algorithm of 7 × 7 templates, is then expanded with 7 × 7 templates
Recover artwork, then deduct image of the artwork of the recovery after 3 × 3 erosion algorithms using the artwork recovered to obtain edge graph
Picture.
The finger two-dimensional coordinate positioning is swept using the edge contour image of whole handss in first and second images of gestures of search
Retouch the angle in hand contour curve, determine angle be acute angle and direction vertically upward for pad of finger, orient finger
Each finger tip position.
The edge image of the two-dimensional localization coordinate and second gesture image of the edge image by first gesture image
Two-dimensional localization Coordinate Conversion is realized into three-dimensional localization coordinate by binocular vision algorithm.
Compared with prior art, two infrared cameras be present invention employs two-dimentional handss are solved obtaining images of gestures
Gesture identification does not have depth information to cause gesture identification inaccurate；Infrarede emitting diode illumination solves asking for visible ray interference
Inscribe, and the camera lens of infrared camera is arranged upward, can reduce the interference of the image background absorbed, unlike prior art is absorbed
Images of gestures often disturbed by human body background, the simple identification for calculating and realizing complicated gesture can be passed through, improved
The accuracy rate of identification.
Description of the drawings
Fig. 1 is the theory structure block diagram of three-dimension gesture identifying device of the present invention.
Fig. 2 is three-dimension gesture identifying device specific embodiment of the present invention（Two infrared cameras and infraluminescence are only shown
Diode）Perspective structure schematic diagram.
Fig. 3 is that the viewing angle constructions of the infraluminescence photographic head of three-dimension gesture identifying device specific embodiment of the present invention are illustrated
Figure.
Fig. 4 is the workflow diagram of three-dimensional gesture recognition method of the present invention.
Fig. 5 is the finger locating structural representation of indication two dimension finger coordinate setting in Fig. 4.
Specific embodiment
To make the objects, technical solutions and advantages of the present invention clearer, clear and definite, develop simultaneously embodiment pair referring to the drawings
The present invention is further described.It should be appreciated that specific embodiment described herein is only to explain the present invention, and without
It is of the invention in limiting.
As shown in figure 1, three-dimension gesture identifying device of the present invention includes two infrared cameras 21,22, image acquisition units
20th, infrarede emitting diode 41, infrarede emitting diode driver element 40, arithmetic element 30 and gesture identification unit 50.
The image acquisition units 20 illustrated in Fig. 1 are electrically connected with two infrared cameras 21,22, two pole of the infraluminescence
Pipe driver element 40 is electrically connected with infrarede emitting diode 41 to drive infrarede emitting diode to light, the arithmetic element 30 with
Image acquisition units 20, infrarede emitting diode driver element 40, gesture identification unit 50 are electrically connected；Described image collecting unit
20 gather images of gestures data by two infrared cameras 21,22, and images of gestures data is activation is entered to arithmetic element 30
Row is processed, and arithmetic element 30 obtains images of gestures data and sends data to gesture identification unit 50 after carrying out three-dimensional coordinate positioning
Carry out pattern match and realize that three-dimension gesture is recognized.When two infrared cameras carry out image capture, the arithmetic element can be with root
According to ambient light, infrarede emitting diode 41 is controlled by infrarede emitting diode driver element 40 and is lighted as two infrared photographies
The images of gestures of head intake provides bias light.
Specifically, the product of three-dimension gesture identifying device of the present invention is set to a box body, as shown in Fig. 2 will be above-mentioned
The building block of device is positioned in the box body 10, and Fig. 2 only illustrates the set location of two photographic head and infrarede emitting diode,
Remaining part can be integrally provided on PCB, not be limited herein, only to illustrate the dress when specific product is formed
The gesture identification principle put.Two infrared cameras 21,22 are horizontally set in three-dimension gesture identifying device 10, and its camera lens is upward
Arrange.Fig. 3 shows the structure that two infrared cameras are arranged, and the distance in two infrared camera horizontal directions is 6cm, is accorded with
The reasonable placement scope of binocular vision as unification, in figure, the intersecting region 23 in two infrared camera visual angles is object run area
Domain, i.e. operation object --- the operating area residing for handss, in the range of this object run region, it is possible to achieve device opponent
The identification of gesture, according to the distance in two infrared cameras 21,22 horizontal directions, the intersecting angle in two infrared camera visual angles is
60°.During work, the top that handss are placed on box body 10 is operated, can so avoid photographic head from obtaining the ambient interferences of image,
Improve computational efficiency.
It is the process chart of three-dimensional gesture recognition method embodiment of the present invention shown in Fig. 4, the three-dimensional gesture recognition method
In gesture extract be that, based on binocular vision technology, its method and step is as follows：
S100, left and right two infrared cameras is horizontally disposed with, and cam lens is upward；
S200, described image collecting unit are obtained on two infrared cameras respectively by left and right two infrared cameras
First gesture image, second gesture image in square angular field of view is sent to arithmetic element；First, second gesture described here
Image refers to the images of gestures absorbed by left or right photographic head, due to being that two infrared cameras are absorbed respectively, uses herein
First, second making a distinction.Above described two infrared cameras, angular field of view is the intersecting 60 ° of folders in two infrared camera visual angles
Angle range, the scope are the object run region in Fig. 3, and during operation, handss work in the object run region.
First gesture image, second gesture image are carried out image binaryzation process by S300, arithmetic element respectively；Described two
Value algorithm is realized using common Threshold Segmentation Algorithm.
First gesture image after binary conversion treatment, second gesture image are carried out edge and are carried by S400, arithmetic element respectively
Take algorithm and obtain the edge image of first gesture image, the edge image of second gesture image；The Boundary extracting algorithm is first adopted
Spuious point is removed with the erosion algorithm of 7 × 7 templates, then recovers artwork with the expansion of 7 × 7 templates, then deducted using artwork is recovered
The figure of the recovery artwork after 3 × 3 erosion algorithms obtains edge image.
S500, arithmetic element realize two according to the edge image of first gesture image, the edge image of second gesture image
Dimension finger coordinate setting；Edge contour of the finger two-dimensional coordinate positioning using whole handss in first and second images of gestures of search
Image, scans the angle in hand contour curve, determine angle be acute angle and direction vertically upward for pad of finger, positioning
Go out each finger tip position of finger.Fig. 5 shows the profile diagram of the edge image of first or second gesture image, such as Fig. 5
Shown, arithmetic element scans the angle in hand contour curve by searching for the edge contour image of whole handss, A such as in figure,
The angle α formed by B, C point, can orient the finger tip position of finger, to obtain two-dimentional finger coordinate.
S600, arithmetic element is by the two-dimensional localization coordinate of the two-dimensional localization coordinate and second gesture image of first gesture image
Associative operation is converted into three-dimensional localization coordinate, carries out matching knowledge with the gesture model information for being pre-stored within gesture identification unit
Not.Specifically, described two-dimensional localization coordinate be according to first or second gesture image edge image obtaining the wheel of handss
Wide coordinate, by the two-dimensional localization Coordinate Conversion of the two-dimensional localization coordinate of first gesture image and second gesture image into three-dimensional localization
Coordinate realizes that by binocular vision method binocular vision algorithm is prior art, is not repeated herein.
Interference of the visible ray to image can be solved using apparatus of the present invention and method, handss are operated above device
The interference of background can be effectively reduced, unlike the gesture identifying device of prior art, human body can be to intake before photographic head
Image has interference, makes calculating process be also contemplated that the interference of human body, solves the complexity of binocular vision calculating, improves gesture
The reliability of identification.
It should be appreciated that the foregoing is only presently preferred embodiments of the present invention, it e insufficient to limit the present invention's
Technical scheme, for those of ordinary skills, within the spirit and principles in the present invention, can add according to the above description
To increase and decrease, replace, convert or improve, and the technical scheme after all these increases and decreases, replacement, conversion or improvement, should all belong to this
The protection domain of invention claims.
Claims (8)
1. a kind of three-dimension gesture identifying device, for being identified to gesture, it is characterised in that including camera unit, image
Collecting unit, infrarede emitting diode unit, infrarede emitting diode driver element, arithmetic element and gesture identification unit；Institute
Stating camera unit includes two infrared cameras；The infrarede emitting diode unit includes infrarede emitting diode；It is described
The camera lens of two infrared cameras is arranged upward, and described image collecting unit is electrically connected with two infrared cameras, described infrared
Light emitting diode driver element is electrically connected with infrarede emitting diode to drive infrarede emitting diode to light, the arithmetic element
Electrically connect with image acquisition units, infrarede emitting diode driver element, gesture identification unit；
Described image collecting unit, is regarded above two infrared cameras for being obtained by left and right two infrared cameras respectively
First gesture image, second gesture image in angle range is sent to arithmetic element；
The arithmetic element, for first gesture image, second gesture image are carried out image binaryzation process respectively；By two-value
First gesture image, second gesture image after change process carries out the edge that Boundary extracting algorithm obtains first gesture image respectively
The edge image of image, second gesture image；Edge image, the edge image of second gesture image according to first gesture image
Realize that finger two-dimensional coordinate is positioned；By the side of the two-dimensional localization coordinate and second gesture image of the edge image of first gesture image
The two-dimensional localization Coordinate Conversion of edge image is believed into three-dimensional localization coordinate with the gesture model being pre-stored within gesture identification unit
Breath carries out match cognization；
The arithmetic element, is additionally operable to control the images of gestures that infrarede emitting diode lights as the intake of two infrared cameras and carries
For bias light.
2. three-dimension gesture identifying device according to claim 1, it is characterised in that between described two infrared cameras
Horizontal range is 6cm.
3. three-dimension gesture identifying device according to claim 2, it is characterised in that the visual angle of described two infrared cameras
For 60 °.
4. three-dimension gesture identifying device according to claim 1, it is characterised in that the infrared diode is at least
It is individual.
5. a kind of three-dimension gesture identifying device based on described in any one of Claims 1-4 realizes three-dimensional gesture recognition method, its
It is characterised by, methods described comprises the steps：
Left and right two infrared cameras are horizontally disposed with, and cam lens are upward；
Described image collecting unit obtains angular field of view above two infrared cameras respectively by left and right two infrared cameras
Interior first gesture image, second gesture image are sent to arithmetic element；
First gesture image, second gesture image are carried out image binaryzation process by arithmetic element respectively；
First gesture image after binary conversion treatment, second gesture image are carried out Boundary extracting algorithm and are obtained by arithmetic element respectively
The edge image of first gesture image, the edge image of second gesture image；The Boundary extracting algorithm first adopts 7 × 7 templates
Erosion algorithm remove spuious point, then recover artwork with the expansion of 7 × 7 templates, then deduct the recovery artwork using artwork is recovered
Figure after 3 × 3 erosion algorithms obtains edge image；
Arithmetic element realizes finger two-dimensional coordinate according to the edge image of first gesture image, the edge image of second gesture image
Positioning；The finger two-dimensional coordinate positioning scans handss using the edge contour image of whole handss in first and second images of gestures of search
Angle in contouring curve, determine angle be acute angle and direction vertically upward for pad of finger, orient the every of finger
Individual finger tip position；
Arithmetic element is by the edge image of the two-dimensional localization coordinate and second gesture image of the edge image of first gesture image
Two-dimensional localization Coordinate Conversion is carried out with the gesture model information being pre-stored within gesture identification unit into three-dimensional localization coordinate
With identification.
6. three-dimensional gesture recognition method according to claim 5, it is characterised in that regard above described two infrared cameras
Angle range is that 60 ° of angular ranges are intersected at two infrared camera visual angles.
7. three-dimensional gesture recognition method according to claim 5, it is characterised in that the Binarization methods adopt common threshold
Value segmentation is realized.
8. three-dimensional gesture recognition method according to claim 5, it is characterised in that the edge by first gesture image
The two-dimensional localization coordinate of image with the two-dimensional localization Coordinate Conversion of the edge image of second gesture image into three-dimensional localization coordinate is
Realized by binocular vision algorithm.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210565201.7A CN103514437B (en) | 2012-12-24 | 2012-12-24 | A kind of three-dimension gesture identifying device and three-dimensional gesture recognition method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210565201.7A CN103514437B (en) | 2012-12-24 | 2012-12-24 | A kind of three-dimension gesture identifying device and three-dimensional gesture recognition method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103514437A CN103514437A (en) | 2014-01-15 |
| CN103514437B true CN103514437B (en) | 2017-04-05 |
Family
ID=49897134
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201210565201.7A Active CN103514437B (en) | 2012-12-24 | 2012-12-24 | A kind of three-dimension gesture identifying device and three-dimensional gesture recognition method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103514437B (en) |
Families Citing this family (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104484679B (en) * | 2014-09-17 | 2017-12-22 | 北京邮电大学 | Non- standard rifle shooting warhead mark image automatic identifying method |
| CN106491071A (en) * | 2015-09-06 | 2017-03-15 | 中兴通讯股份有限公司 | A kind of method for giving a test of one's eyesight and terminal |
| CN105892637A (en) * | 2015-11-25 | 2016-08-24 | 乐视致新电子科技（天津）有限公司 | Gesture identification method and virtual reality display output device |
| CN106211470B (en) * | 2016-08-24 | 2018-02-13 | 横店集团得邦照明股份有限公司 | A kind of LED dimming control systems and its implementation based on 3D gesture controls |
| CN106443651A (en) * | 2016-09-30 | 2017-02-22 | 乐视控股（北京）有限公司 | Virtual reality device, and spatial positioning system and method |
| TWI636395B (en) * | 2016-11-10 | 2018-09-21 | 財團法人金屬工業研究發展中心 | Gesture operation method and system based on depth value |
| CN106547359A (en) * | 2016-11-30 | 2017-03-29 | 珠海迈科智能科技股份有限公司 | High voltage power supply with gesture identification function and its recognition methodss |
| CN107896274B (en) * | 2017-10-27 | 2020-08-21 | Oppo广东移动通信有限公司 | Infrared transmitter control method, terminal and computer-readable storage medium |
| CN108363482A (en) * | 2018-01-11 | 2018-08-03 | 江苏四点灵机器人有限公司 | A method of the three-dimension gesture based on binocular structure light controls smart television |
| CN114390760B (en) * | 2022-01-20 | 2023-11-21 | 北方工业大学 | Light control method and system |
| CN116363795A (en) * | 2023-03-22 | 2023-06-30 | 湖南云数信息科技有限公司 | Vending cabinet, management method and storage medium |
Family Cites Families (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101510121A (en) * | 2009-03-12 | 2009-08-19 | 重庆大学 | Interface roaming operation method and apparatus based on gesture identification |
| TWI489317B (en) * | 2009-12-10 | 2015-06-21 | Tatung Co | Method and system for operating electric apparatus |
| CN101847057A (en) * | 2010-06-01 | 2010-09-29 | 郭小卫 | Method for touchpad to acquire input information |
| CN102156859B (en) * | 2011-04-21 | 2012-10-03 | 刘津甦 | Perception method of hand posture and spatial position |
| CN102184008A (en) * | 2011-05-03 | 2011-09-14 | 北京天盛世纪科技发展有限公司 | Interactive projection system and method |
| CN102750046B (en) * | 2012-05-30 | 2015-09-09 | 浙江大学 | Based on quick three-dimensional display interactive device and the method for the infrared identification of double structure light |
| CN102799271A (en) * | 2012-07-02 | 2012-11-28 | Tcl集团股份有限公司 | Method and system for identifying interactive commands based on human hand gestures |
-
2012
- 2012-12-24 CN CN201210565201.7A patent/CN103514437B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN103514437A (en) | 2014-01-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103514437B (en) | A kind of three-dimension gesture identifying device and three-dimensional gesture recognition method | |
| US9898651B2 (en) | Upper-body skeleton extraction from depth maps | |
| CN103941866B (en) | Three-dimensional gesture recognizing method based on Kinect depth image | |
| CN103226387B (en) | Video fingertip localization method based on Kinect | |
| CN104463146B (en) | Posture identification method and device based on near-infrared TOF camera depth information | |
| CN107038424B (en) | A gesture recognition method | |
| CN103984928B (en) | Finger gesture recognition methods based on depth image | |
| CN110555412B (en) | End-to-end human body gesture recognition method based on combination of RGB and point cloud | |
| CN101477631B (en) | Method, device and human-computer interaction system for extracting target from image | |
| CN101776952B (en) | Novel interactive projection system | |
| CN103714322A (en) | Real-time gesture recognition method and device | |
| CN104574365B (en) | Obstacle detector and method | |
| CN108399361A (en) | A kind of pedestrian detection method based on convolutional neural networks CNN and semantic segmentation | |
| CN103198304A (en) | Palm print extraction and identification method | |
| CN103927016A (en) | Real-time three-dimensional double-hand gesture recognition method and system based on binocular vision | |
| CN102323868A (en) | Human-computer multi-point interaction method and device based on infrared image | |
| CN105046219A (en) | Face identification system | |
| CN106503651B (en) | A kind of extracting method and system of images of gestures | |
| CN103500010B (en) | A kind of video fingertip localization method | |
| Xu et al. | Integrated approach of skin-color detection and depth information for hand and face localization | |
| CN103426008B (en) | Visual human hand tracking and system based on online machine learning | |
| CN106650628B (en) | Fingertip detection method based on three-dimensional K curvature | |
| CN103440035A (en) | Gesture recognition system in three-dimensional space and recognition method thereof | |
| CN110607405A (en) | Leather inner contour recognition and cutting device and method based on machine vision industrial application | |
| CN106484108A (en) | Chinese characters recognition method based on double vision point gesture identification |
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 |