Plug-in element stitch vision positioning device based on binocular vision
Technical field
This utility model relates to the use of the non-contact method of binocular vision and positions the stitch of plug-in element, thus right
Plug-in element positions, a kind of plug-in element stitch vision positioning device based on binocular vision.
Background technology
Along with the market development of electronic product, the production efficiency of electronic product needs to improve constantly, production automatic
Change degree needs to improve constantly.A part of electronic product in the market, because the requirement to volume is less stringent, in order to
Control cost, extensive application plug-in element.Along with improving constantly of electronic product demand, the efficiency of artificial plug-in unit and quality are
Through current production needs cannot be met, need to be replaced the process of artificial plug-in unit, automatization's plug-in unit by automatization's plug-in component equipment
The core technology of equipment is a kind of element localization method accurately and quickly.
Position plug-in unit electronic component currently with machine vision, have by side laser to stitch polishing, phase bottom utilization
Machine 100 gathers the refraction to light of the stitch end to position stitch, as shown in Figure 1.This method mechanical hand 101 clamps to be detected
Element 102, laser line light source 103 is from the side polishing of element to be detected, and the light of line source shines stitch bottom, through stitch
Bottom chamfering reflection light enters camera lens, imaging in bottom camera.If but component body to be detected is white, body
The reflective location with printing opacity severe jamming stitch, is limited bigger by body color and ambient light;In addition due to pin
The discordance of foot portion chamfer angle technique, the mass discrepancy of these chamferings can affect the accuracy of location.
Currently with machine vision positioning electronic element, utilize the phase of opposite side also by side directional light 103 polishing
The projection of machine 100 acquisition elements body all directions determines position and the attitude of element, as shown in Figure 2,3.This method is to use
Sucker 101 holds electronic component 102, and through directional light 103 polishing of element side, component body to be detected shelters from a part of light
Line, the light at other positions enters the camera lens of the opposite side at element to be detected, forms the projection picture of component body at camera.
But the method can only pass through sucker extracting elements, since it is desired that determined position and the appearance of element by the projection of component body
State, so manipulator clamping element to be positioned can not be passed through, so the rough electronic component of some upper surfaces cannot be applied to;
And need to coordinate the projection picture of rotary motion acquisition elements body multiple directions, the most computationally intensive, and this location side
Method needs element to be detected to coordinate rotary motion to gather the projection picture of multiple directions in the detecting position pause regular hour, if
Using automatization's plug-in component equipment, this localization method can make plug-in unit efficiency be substantially reduced.A part of body shape is not advised
Then or the highest plug-in element of its stitch opposing body's positional precision, this localization method by body, the mistake of stitch location
Difference is bigger, is likely to result in stitch and cannot insert circuit board.
Summary of the invention
The purpose of this utility model is the shortcoming overcoming above-mentioned prior art, it is provided that plug-in element based on binocular vision
Stitch vision positioning device, by the projection picture of acquisition elements stitch, can the most stably detect all stitch, not treated
Detecting element body color and the impact of stitch end-beveling quality;And by stitch is directly positioned, overcome body
The precision shortcoming of projective iteration, and need to gather the inefficient shortcoming of different directions image by rotating element.
For achieving the above object, this utility model adopts the following technical scheme that
Plug-in element stitch vision positioning device based on binocular vision, this device includes two cameras, a light source, one
Individual linear motion device, a cylinder clamping device clamping element to be positioned;Wherein:
Two cameras are positioned at the side of the stitch of element to be positioned, and light source is positioned at the opposite side of element to be positioned, and light leads to
Crossing the pin area of element to be positioned, the outline projection of target stitch is at camera imaging, the stitch projection of two collected by cameras
As may determine that the stitch position on two camera optical axis place planes;
Cylinder gripper mechanism grips element to be positioned, by linear motion device control element to be positioned by two cameras and
Region between light source, makes two cameras can collect target stitch projection by the speed controlling linear motion device
Picture, is determined the position of element to be positioned by the displacement transducer of linear motion device, and return signal is given and controlled computer control camera
Take pictures and light source switch.
Such scheme is further: described cylinder clamping device is bridged by rotational motion mechanism with linear motion device.
Such scheme is further: described linear motion device is motor-screw mandrel straight-line motion mechanism, includes servo electricity
Machine, screw mandrel, feed screw nut, feed screw nut is sleeved on screw mandrel, and driven by servomotor screw mandrel rotates, and feed screw nut is axial along screw mandrel
Moving back and forth, feed screw nut connects rotational motion mechanism, and rotational motion mechanism reconnects cylinder clamping device.
Such scheme is further: described computer is communicatively coupled motion controller, image pick-up card by bus, is taken the photograph
As controller, light source controller, computer is provided with image processing module, CPU, memorizer, man-machine interface,
Described motion controller is connected with described linear motion device;Said two camera and described camera controller
Connecting, light source is connected with light source controller, and described camera controller is for realizing the image acquisition of said two camera, described light
Source controller, for realizing the on-off control of described light source, is used for coordinating collected by camera image;The signal of said two camera is defeated
Going out end to be connected respectively with described image pick-up card, described image pick-up card is for giving institute by the data image signal of collected by camera
State CPU to process or preserve to described memorizer;The displacement transducer of described linear motion device is connected with I/O mouth, feeds back to be checked
The position of survey element, to CPU, is used for controlling camera and light source works.
This utility model compared with prior art, has the following advantages and beneficial effect:
(1) this vision positioning method can detect element stitch to be detected by the cameras that two optical axises are vertical with stitch
Position in a horizontal plane and the attitude of element.
(2) this vision positioning method combines straight-line motion mechanism so that camera can take electronic component fortune at gripper of manipulator
The projection picture of acquisition elements stitch during Dong, it is not necessary to by rotating the image information obtaining all angles, Ke Yi great
The big detection and localization time saving plug-in element.
(3) this vision positioning method is by the way of backlight polishing, utilizes the projection of stitch as treating detecting element
Stitch positions, it is to avoid component body printing opacity to be detected and reflective and stitch end-beveling quality are to stitch locating effect
Impact.
(4) this vision positioning method carrys out setting element by the image of direct acquisition elements stitch, effectively eliminates and passes through
The image of component body carrys out the error of setting element position, substantially increases the precision of element location.
(5) this utility model can be used for the high-speed, high precision automatic optics inspection of plug-in part technology industry.
Accompanying drawing illustrates:
Below in conjunction with the accompanying drawings and this utility model is further described by case study on implementation:
Fig. 1 is existing side laser polishing stitch positioner sketch;
Fig. 2 is existing electronic element positioner sketch;
Fig. 3 be Fig. 2 look up sketch;
Fig. 4 is this utility model component placement device front view;
Fig. 5 is this utility model component placement device top view;
Fig. 6 is common lens pin-hole imaging model;
Fig. 7 is double common lens camera coordinates system;
Fig. 8 is double camera coordinate system and vertical coordinate system;
Fig. 9 is vertical coordinate system and world coordinate system;
Figure 10 is two stitch location schematic diagrams;
Figure 11 is doubly telecentric lens camera coordinate system;
Figure 12 this utility model detecting system structure block diagram.
Detailed description of the invention:
Below in conjunction with embodiment and accompanying drawing, this utility model is described in further detail, but enforcement of the present utility model
Mode is not limited to this.
Refering to shown in Fig. 4 ~ 12, the plug-in element stitch based on binocular vision that this utility model provides for said method
Vision positioning device includes two cameras 6,7, a light source 8, a linear motion device, the gas clamping element to be positioned
Cylinder clamping device 4;Wherein:
Two cameras 6,7 are positioned at the side of the stitch of element 5 to be positioned, and light source 8 is positioned at the opposite side of element 5 to be positioned,
The light pin area by element to be positioned, the outline projection of target stitch 51 is at camera imaging, the pin of two collected by cameras
Foot projection picture may determine that the stitch position on two camera optical axis place planes;
Cylinder clamping device 4 clamps element to be positioned, linear motion device control element to be positioned by two cameras
And the region between light source, make two cameras can collect target stitch 51 by the speed controlling linear motion device and throw
Shadow image, is determined the position of element to be positioned by the displacement transducer of linear motion device, and return signal is given and controlled computer control
Camera is taken pictures and light source switch.
Embodiment 1:
As shown in Figure 4,5, this device includes camera 6, camera 7, light source 8, cylinder clamping device 4 and linear motion device.
Cylinder clamping device 4 is bridged by rotational motion mechanism 3 with linear motion device.Described linear motion device is motor-screw mandrel
Straight-line motion mechanism, includes servomotor 11, screw mandrel 9, feed screw nut 2, and screw mandrel 9 is arranged on support 1 by rolling bearing 10
On, feed screw nut 2 is sleeved on screw mandrel 9, and servomotor 11 drives screw mandrel 9 to rotate, and feed screw nut 2 moves along screw mandrel axial reciprocating
Dynamic, feed screw nut 2 connects rotational motion mechanism 3, and rotational motion mechanism 3 reconnects cylinder clamping device 4.As shown in figure 12, institute
Stating computer and be communicatively coupled motion controller, image pick-up card, camera controller, light source controller by bus, computer sets
Image processing module, CPU, memorizer, man-machine interface, described motion controller is had to be connected with described linear motion device;
Said two camera is connected with described camera controller, and light source is connected with light source controller, and described camera controller is used for realizing
The image acquisition of said two direction camera, described light source controller, for realizing the on-off control of described light source, is used for coordinating
Collected by camera image;The signal output part of said two direction camera is connected respectively with described image pick-up card, and described image is adopted
Truck processes for the data image signal of collected by camera is given described CPU or preserves to described memorizer;Described straight line is transported
The displacement transducer of the servomotor 11 on dynamic device is connected with I/O mouth, feeds back the position of element to be detected to CPU, is used for controlling
Camera processed and light source works.
Camera and light source lay respectively at the both sides of element to be positioned, and light is perpendicular to the direction of motion of element to be detected, phase
Machine 6 and the angled distribution of camera 7.The optical axis of the camera lens of camera 6 and camera 7 is all perpendicular to the side of element stitch to be positioned
To, and the optical axis of the camera lens of camera 6 and camera 7 is positioned at same level, s-O-t plane as shown in Figure 7.
Cylinder clamping device 4 clamps element to be detected, first controls rotational motion mechanism 3 and rotates to an angle so that camera 6
The target stitch gathered with camera 7 is without the overlap of other stitch.Then unit to be detected is driven by motor-screw mandrel straight-line motion mechanism
Region between part straight-line pass camera and light source.The position of straight-line motion mechanism is obtained by the displacement transducer of servomotor,
This position signalling is fed back to industrial PC, controls camera 6 with this position signalling and camera 7 is taken pictures, and switch as light source 8
Signal source, i.e. when original paper to be positioned moves to specify position, light source 8 opens, and camera 6 and camera 7 are taken pictures, then motor-
Screw mandrel straight-line motion mechanism moves forward, and after motion certain distance, the second target stitch of element to be positioned arrives two
After the shot region of individual camera, the signal of the displacement transducer of servomotor feed back to industrial PC and control camera 6 and camera 7
Take pictures, be then shut off light source 8.
The principle of this image acquisition device image: light passes pin area, stitch from the side of element to be detected
Sheltering from a part of light, in camera Shang Cheng black view field, remaining printing opacity position becomes white portion on camera, in conjunction with phase
In machine 6 and camera 7, element stitch can be positioned by the projection picture of stitch.
As shown in figure 12, above-mentioned plug-in element stitch vision positioning device based on binocular vision is used.Computer is by total
Line is communicatively coupled motion controller, image pick-up card, camera controller, light source controller, wherein: camera controller is respectively
It is connected with camera 6 and camera 7, controls camera 6 and camera 7 image acquisition;Light source controller is connected with light source 8, controls light source 8.
The signal of image pick-up card and camera 6 and camera 7 exports and is connected, the data image signal of collected by camera is given CPU process or
It is saved in memorizer.Motion controller is connected with described linear motion device, is used for controlling linear motion.Linear motion device
Servomotor displacement transducer be connected with I/O mouth, feed back the position of element to be detected to CPU, camera 6, camera 7 and light source 8
Signals collecting image according to sensor, image acquisition process is completed by procedure auto-control, the image of collection and result
It is saved in memorizer, for determining the position relationship of two stitch.
Computer carries out binaryzation to the image gathered, and is then analyzed bianry image, finds out the stitch needing location,
Then coordinate system is changed, finally give element stitch position in world coordinate system and the attitude of element.
Embodiment 2:
Based on embodiment 1, this binocular vision component position system also includes stitch location and coordinate transformation method, Yi Jiyuan
The determination method of part attitude:
Because the parameter of each camera and camera lens is different, thus location stitch before need calibration for cameras in and outward
Parameter, element stitch localization method of the present utility model needs the intrinsic parameter demarcated to include the core of lens camera focal length and camera
The size of sheet unit picture element.Demarcation except camera intrinsic parameter, in addition it is also necessary to the outer parameter of calibration for cameras, determines two camera optical axises
Angle, and determine position in world coordinate system, the lens optical center by demarcation, determine two cameras the most again
The distance of the initial point O of optical center distance double camera coordinate system s-O-t.
As shown in Figure 6, common lens is national forest park in Xiaokeng, the camera lens of combining camera 6 and two apertures of the camera lens of camera 7
Imaging model, determines the projection of stitch position in the picture, i.e. stitch projection picture position in image coordinate system.Pass through
National forest park in Xiaokeng can converse the stitch to be positioned position coordinates P(s in the camera coordinates system that two cameras are constituted, t).
Then stitch position is transformed in orthogonal coordinate system x-O-y of coordinate axes as shown in Figure 8, it is assumed that this coordinate
The x-axis of system is parallel with the s axle of camera coordinates system, and initial point overlaps with s-O-t coordinate system.By camera coordinates system, (s t) is transformed into x-
The formula of O-y coordinate system is:()
()
The k axle of hypothetical world coordinate system k-O'-w is parallel with member motion direction, is then turned in x-O-y coordinate system by P point
In coordinate system k-O'-w changed to, as it is shown in figure 9, obtain stitch P expression P(k in k-O'-w coordinate system after Zhuan Huan, w), turn
Changing formula is:
So far, only determine stitch position in world coordinate system, remember that first stitch positioned is alive
Position coordinates under boundary's coordinate system is (k1, w1), and next step to determine the position of element geometric center by the position of two stitch
Put the attitude with element.Mechanical hand continues to clamp member motion to be positioned, records this distance for d, and first stitch motion puts in place
Put P'(k1-d, w1), as shown in Figure 10, the most again second stitch is taken pictures, determined by method above the most again
The position Q (k2, w2) of second stitch.Then by the position calculation of the two stitch go out element geometric center position and
The attitude of element.
Embodiment 3:
Telecentric lens binocular vision plug-in element localization method based on embodiment 1 shown in Figure 11:
Common lens image-forming principle is national forest park in Xiaokeng based on Fig. 6, needs the inside and outside parameter aperture by camera
Imaging model just can obtain stitch coordinate in double camera coordinate s-O-t through more complicated calculating.If at this binocular
Using two telecentric lens in vision plug-in element alignment system, telecentric lens is the imaging model of a kind of directional light, by far
Heart camera lens obtains the projection picture of stitch, and the calculating of telecentric lens imaging model is simpler, can simplify image further and sit
Mark system is transformed into the process of double camera coordinate system s-O-t.And telecentric lens can effectively reduce the distortion of lens imaging, permissible
Simplify calculating process further and improve the positioning precision of binocular vision plug-in element alignment system.
Embodiment 4:
Based on embodiment 1, this image collecting device needs to ensure that two stitch may detect that:
Need to rotate a certain angle so that the element two that two cameras obtain after mechanical hand captures element to be positioned
The projection picture of target stitch does not has other stitch overlapping, and concrete angle needs the stitch according to element to be positioned to be distributed and stitch
Width determines.Rotational motion mechanism is contemplated to ensure that two target stitch do not have other stitch overlapping, thus ensures positioning accurate
Degree.
After cylinder gripper mechanism grips element rotates to an angle, then moving along a straight line, target stitch moves to examine
Surveying region, the region that i.e. two cameras can photograph, as shown in the green overcast region in Fig. 5, then light source is opened, two
Camera is taken pictures to first aim stitch.Then motor-screw mandrel straight-line motion mechanism control mechanical hand continues to move along a straight line, directly
Second target stitch to element to be positioned arrives detectable region, and two cameras take pictures to respectively second target stitch,
Ensure that two target stitch can detect to obtain.