Background technology
In teaching, meeting and issue field, people need write and explain the content of oneself usually, traditional media is blackboard and paper, but development along with infotech, people progressively rely on computing machine by the large screen display content, present alternately basically by means of mouse and keyboard, but also can't the drawing, write intuitively by referring to as blackboard alternately of this explanation formula.At above-mentioned deficiency, use the electronic whiteboard of touching technique progressively to obtain promoting, it obtains the motion track and the touch information of object in the view field by auxiliary electronic device, and connects computer system realization man-machine interaction.
Electronic whiteboard can be divided into two classes by touching technique: initiatively touch and passive touch, touch system initiatively needs special device to remove to contact touch-surface, and the most frequently used interactive device is the interaction pen that special power supply is powered; Passive touch system does not then need special contact deactivation touch-surface, and the user can finish interworking with finger or other objects fully.
From the development trend of whole electronic whiteboard, passive touch type electronic blank has been tending towards main flow, adopts the Detection Techniques of shooting scan-type to have following several:
First kind, use is installed in the outer video camera of blank whole blank is taken in, when light pen is write on blank, camera perception luminous point move the position that obtains light pen, its shortcoming is that camera need be installed in the blank outside, when the people write in front, the luminous point of light pen can be blocked.
Second kind, adopt a plurality of line array CCDs that are installed on the blank that light pen is surveyed, by judging that luminous point carries out compute location in the position of each thread CCD, its shortcoming is to need light pen.
The third, a plurality of video cameras that employing is installed on the blank scan surveying object, carry out coordinate setting by the characteristic quantity of judgment object in each video camera, and this mode still is more advanced, but handle comparatively complexity, the installation to video camera has very high requirement in addition.
Summary of the invention
The present invention will solve the shortcoming of above-mentioned prior art, a kind of electric whiteboard system and detection method based on image detection is provided, and it obtains touch information indirectly by detecting auxiliary blank block diagram picture, it allows finger or object directly to touch on blank, need not to use sign pen.
The present invention solves the technical scheme that its technical matters adopts: this electric whiteboard system based on image detection, mainly comprise blank, left and right sides corner location on blank is placed with the first image shooting device and the second image shooting device respectively, wherein the first image shooting device is used for the second opposite side moulding and the 3rd frame bar carries out image acquisition, the second image shooting device is used for the second opposite side moulding and the first frame bar carries out image acquisition, and the first frame bar, the second frame bar and the 3rd frame bar are provided with black and white at the medial surface of blank and replace the border; The central optical axis of the first image shooting device and the second image shooting device is angled with its both sides frame respectively, and the height of central optical axis is right against black and white and replaces the border; The first image shooting device is connected with primary processor respectively with the second image shooting device.
The described first image shooting device and the second image shooting device constitute by camera and graphics processing unit, and the central optical axis of two cameras becomes 45 degree with its both sides frame respectively.
Described camera is made up of greater than the sphere camera lens of 90 degree and the CMOS color image sensor of high-resolution the horizontal field of view angle.
Described primary processor is connected with computing machine by the USB device.
The detection method of this electronic whiteboard based on image detection of the present invention, step is as follows:
1, the left and right sides corner location on blank is placed the first image shooting device and the second image shooting device respectively, the first frame bar, the second frame bar and the 3rd frame bar are provided with black and white at the medial surface of blank and replace the border, the central optical axis of the first image shooting device and the second image shooting device is angled with its both sides frame respectively, adjust the height of central optical axis, make it be right against black and white and replace the border;
2, the first image shooting device carries out image acquisition to the second opposite side moulding and the 3rd frame bar, and the second image shooting device carries out image acquisition to the second opposite side moulding and the first frame bar;
3, replacing the border by black and white is the touch plane that GTG obtains reference, when object enters the touch area, GTG edge in the frame bar will cause destruction, draw the crossing angle of object between two camera heads by breakdown point, and obtain the position of object in the blank zone by the triangulation calculation method; The gray scale image of the internal side frame of graphics processing unit carries out edge extracting, and when not having object to touch, the GTG edge contour of this extraction is continuously and collimates; When object touched blank, the frame image was subjected to stopping of object, and its edge contour can produce distortion or fracture, just can calculate the position of object at blank by knowing object at the deformation point that two or more camera heads produces.Image in the first image shooting device and the second image shooting device dialogue edges of boards frame is done rim detection, adopt the benefit of differentiating the frame decorative edging, be that also area planar that its frame GTG edge surrounds is with forming a Petting Area between the blank plane, when the edge is destroyed, mean that also object has entered touch condition.
Under the situation of having used above-mentioned edge contour to detect, the present invention is also auxiliary to have adopted mobile detection to differentiate, image in the first image shooting device and the second image shooting device dialogue edges of boards frame is done to move and is detected, its objective is and improve the accuracy that detects, graphics processing unit is by analyzing front and back two frame differences of frame image, and whether judgment object moves near Petting Area.
In this invention system, the first image shooting device and the second image shooting device are done color identification to the object that moves in the blank, by the color component of object analysis at color image sensor, draw the color of object, and PC writes painted the information notice.Its benefit is the user uses the object of which kind of color, just can write with which kind of color.
The invention has the beneficial effects as follows: the present invention has adopted the intrinsic frame of blank to carry out image detection, is not increasing under the additional electronic devices situation, has realized the interactive information of electronic whiteboard.It, is changed by the characteristic quantity of judging this reference target as the image detection target by the blank frame of introducing, and obtains the touch information of object at electronic whiteboard indirectly, and the invention of native system becomes simply board for writing or drawing, has high practical value.This external system is also further introduced to move and is detected and color identification, and the former improves the accuracy that detects, and the latter provides electronic intelligence color pencil function.
Description of drawings
Fig. 1 is that system of the present invention forms synoptic diagram;
Fig. 2 decorates synoptic diagram for system of the present invention frame bar;
Fig. 3 is the composition synoptic diagram of image shooting device of the present invention;
Fig. 4 is the principle schematic of graphics processing unit of the present invention;
Fig. 5 is the principle schematic of main processor unit of the present invention;
Fig. 6 is a coordinate Calculation synoptic diagram of the present invention;
Fig. 7 is a secondary light source scheme of installation of the present invention;
Description of reference numerals: the first image shooting device, 1, the second image shooting device, 2, the first frame bars 3, the second frame bar, 4, the three frame bars 5, blank 6, primary processor 7, black and white replace border 11, sphere camera lens 20, CMOS color image sensor 21, programmable logic device (PLD) 22, digital signal processor 23, camera 26, graphics processing unit 27, view data snubber assembly 51, picture noise filtration module 52, edge extracting module 53, mobile detection module 54, color average module 55, digital signal module 56, brightness data 60, chroma data 61, tracing analysis unit 62, variance analysis unit 63, color analysis unit 64, single-chip microcomputer 70, EPROM 71, USB module 80, testee 100, infrared LED 101, pcb board 110.
Embodiment
The invention will be further described below in conjunction with drawings and Examples:
As shown in Figure 1, this electric whiteboard system based on image detection mainly comprises blank 6, places the first image shooting device 1 and the second image shooting device 2 (also can more than two) respectively at the left and right sides corner location on the blank 6.The first image shooting device 1 and the second image shooting device 2 constitute (see figure 3) by camera 26 and graphics processing unit 27, camera 26 is made up of greater than the sphere camera lens 20 of 90 degree and the CMOS color image sensor 21 of high-resolution the horizontal field of view angle, the central optical axis of the camera after the installation becomes 45 degree with its both sides frame, 1 pair second frame bar 4 of the first image shooting device and the 3rd frame bar 5 carry out image acquisition, the 2 pairs second frame bars 4 of the second image shooting device and the first frame bar 3 carry out image acquisition, the first frame bar 3, the second frame bar 4 and the 3rd frame bar 5 are decorated black and white at the medial surface of blank 6 and replace border 11 and see Fig. 2 (scuncheon is decorated to GTG and separates), adjust camera with the setting height(from bottom) between blank 6 planes, make its central optical axis also replace the border over against the black and white of frame bar, if clipboard is light, frame is dark, and image detection is carried out on the border that we also can directly use both assemblings to go out.
The detection method of this electronic whiteboard based on image detection of the present invention, step is as follows:
1, the left and right sides corner location on blank 6 is placed the first image shooting device 1 and the second image shooting device 2 respectively, the first frame bar 3, the second frame bar 4 and the 3rd frame bar 5 are provided with black and white at the medial surface of blank 6 and replace border 11, the central optical axis of the first image shooting device 1 and the second image shooting device 2 is angled with its both sides frame respectively, adjust the height of central optical axis, make it be right against black and white and replace border 11;
2,1 pair second opposite side moulding 4 of the first image shooting device and the 3rd frame bar 5 carry out image acquisition, the 2 pairs second opposite side mouldings 4 of the second image shooting device and the first frame bar 3 carry out image acquisition, promptly by two imageing sensors that are installed on the common white edges of boards angle relative frame bar are carried out image scanning;
3, replace border 11 obtains reference for GTG touch plane by black and white, when object enters the touch area, GTG edge in the frame bar will cause destruction, draw the crossing angle of object between two camera heads by breakdown point, and obtain the position of object in the blank zone by the triangulation calculation method.(object or finger cause the frame images acquired to destroy when entering Petting Area, by destroying indirect acquisition touch information)
4, the image in the first image shooting device 1 and the second image shooting device, the 2 dialogue edges of boards frames is done to move and is detected, and 2 pairs of objects that move in the blank of the first image shooting device 1 and the second image shooting device are done color identification.
Graphics processing unit 27 is made of FPGA programmable logic device (PLD) 22 and DSP digital signal processor 23, and FPGA is responsible for the pre-treatment of view data, and DSP is responsible for the aftertreatment of view data.
Primary processor 7 is responsible for the data of acceptance from the first image shooting device 1 and the second image shooting device 2, and positional information after the processing and colouring information send computing machine to by the USB device.Its schematic diagram is seen shown in Figure 5, is made up of the single-chip microcomputer 70 and the EPROM 71 of internal USB module 80, and EPROM 71 stores dispatch from the factory locating information and facility information.
The implementation method of graphics processing unit:
Absorption interval by cutting imageing sensor 21, make the ornamental tape of the inboard frame of its scanning, the view data of its output is admitted to graphics processing unit 27 schematic diagrams and handles as shown in Figure 4, view data snubber assembly 51 in the graphics processing unit 27 is separated into brightness data 60 and chroma data 61 to the view data from imageing sensor 21, wherein brightness data 60 is sent to picture noise filtration module 52 and carries out noise filtering, delivered to edge extracting module 53 simultaneously and mobile detection module 54 is done the image pre-service through the brightness data of noise processed, 61 of chroma data bufferings are admitted to color average module 55 and do the color average and handle.
Below specifically analyze the realization mechanism of above-mentioned image processing module, the length direction of at first establishing picture strip is x, and Width is y:
1) the noise filtering module 52
In realization, we adopt the 2D medium filtering, promptly
g(x,y)=MED{p(x-i,y-j)},(i,j)∈S (1)
Wherein, MED{} for the set in element get intermediate value, p (x y) be grey scale pixel value, g (x, y) be the processing after grey scale pixel value, S is a template window, selects 3 * 3.
2) the edge extracting module 53
In realization, we adopt horizontal edge to extract, promptly
Wherein, (x y) is grey scale pixel value to g, and (x y) is gray-scale value behind the edge extracting to d.
(x y) does binary conversion treatment and refinement, obtains boundary curve y=f (x) to above-mentioned d again for we.
3) mobile detection module 54
In realization, the method that we have adopted the present frame and the initial reference frame that powers on to subtract each other, promptly
m(x,y)=|g
t(x,y)-g
0(x,y)| (3)
Wherein, g
t(x y) is the gradation data of present frame, g
0(x y) is the gradation data of initial reference frame, m (x, y) the comparison frame for exporting.
We again to above-mentioned comparison frame m (x y) does binary conversion treatment, obtain binary picture picture frame n (x, y).
4) color average module 55
In realization, it is average to do chroma data near the boundary curve of our employings to initial extraction, promptly
Wherein, f
0(x) curvilinear function that extracts for initial edge, C
r(x y) is C in the image
rThe pixel value of color difference components, C
b(x y) is C in the image
bThe pixel value of color difference components, K is the chroma sampling window, gets 2,
Mean value for corresponding color difference components.
After the view data process edge extracting module 53 after the noise processed, be sent in the tracing analysis unit 62 in the digital signal module 56 and carry out curve fitting, match is by comparing record coordinate figure x when the match criterion meets with current edge extracting curve to the edge extracting curve that powers on initial.Its match criterion is defined as follows:
|f
t(x)-f
0(x)|>K (5)
Wherein, f
0(x) be the boundary curve function of initial extraction, f
t(x) be the boundary curve of current extraction, K is the match threshold value, gets 2.
View data after the noise processed is through behind the mobile detection module 54, be sent to and carry out the difference differentiation in the variance analysis unit 63 in the digital signal module 56, the template window of its differentiation is slided along the boundary curve of initial extraction, record coordinate figure x when the difference criterion meets.The n when if image difference takes place (x is 1 y), the n when no image difference takes place (x is 0 y), and its difference criterion is defined as follows:
Wherein, (x, the y) binary image data that draws for formula 3, N are the number of pixels in the template window to ADD{}, and S is a template window, selects 5 * 5, (x for element addition in the set, n
0, y
0) be template center's point coordinate, f
0(x) curvilinear function that extracts for initial edge.
Digital signal processing module 56 carries out comprehensive analysis to the two record data again, supposes that the coordinate set by tracing analysis unit 62 records is X
a, be X by the coordinate set of variance analysis unit 63 records
b, the so actual coordinate set X that draws is
X=X
a∩X
b (7)
To coordinate set X border centering point, can obtain the physical location in imageing sensor 21 of object, behind the object space that digital signal processing module 56 obtains, in the color average of this position extraction, convert color value output to by predefined color difference components interval by color analysis unit 64 from color average module 55.
Arrive this, we have obtained positional information and the colouring information of object at the single image shooting device.We connect by serial line interface the information of two image shooting devices again with primary processor 7, the final color value of object is got the average magnitude of both color values, coordinate is then by triangulation calculation in the plane for it, and the calculating synoptic diagram is seen Fig. 6, concrete calculating schematically as follows:
If the first image shooting device, 1 testee, 100 subtended angles are α, the object subtended angle that the second image shooting device 2 measures is β, and the width of blank is L, and the upper left corner of regulation blank is initial point, is downwards y, is to the right x, and then object coordinates is:
x=L×tanβ/(tanα+tanβ), (8)
y=L×tanα×tanβ/(tanα+tanβ)
In order to adapt to the image detection of the internal side frame of energy under any environment, we have also installed auxiliary lighting apparatus on camera head, specifically see synoptic diagram 7, carry out floor light by the 101 pairs of detected inboard frames of infrared LED that are placed on the pcb board 110.
In addition to the implementation, the present invention can also have other embodiments.All employings are equal to the technical scheme of replacement or equivalent transformation formation, all drop on the protection domain of requirement of the present invention.