CN205466320U - Intelligent machine hand based on many camera lenses - Google Patents

Abstract

The utility model discloses an intelligent machine hand based on many camera lenses, include many joints multifunctional machinery hand, be used for to gather to wait to assemble the CCD camera of PCB board image, biological contact device and computer, the CCD camera install many joints multifunctional machinery on hand, biological contact device installs the finger tip at many functions of joint manipulator, CCD camera, biological contact device and many joints multifunctional machinery hand and computer link. The utility model discloses on the basis of two meshes location, biological contact device has added in the manipulator, can carry out the accurate positioning to the assembly target area, improved the rate of accuracy of special -shaped assembly, the utilization be the sensitive material's among the biological contact device pressure - electric output conversion relationship and the manipulator position and the mathematic model of pressure, the utility model discloses a high accuracy and the real -time of dysmorphism components and parts are assembled, have improved the production efficiency of electron assembly industry.

Description

A kind of puma manipulator based on many camera lenses

Technical field

This utility model relates to electronic assemblies field, is specifically related to a kind of puma manipulator based on many camera lenses.

Background technology

At present, in mechano-electronic assembling industry, abnormity components and parts are installed on PCB often rely on flowing water Workman on line carries out hand assembled.Manpower assembling mainly relies on human eye and experience, in work for a long time After cannot ensure production efficiency, and hand assembled is with certain subjectivity, can not maintain every product Equal quality, therefore automatic technology is incorporated into electronic assemblies industry extremely urgent.Automatic technology is taken Production efficiency can be effectively improved for manpower assembling, save human resources, it is ensured that the quality of assembling.

Relevant enterprise has utilized mechanical hand to replace staff to be operated at assembling line, but owing to mechanical hand does not has Someone eyes position, also unlike workman can rely on the sense of touch experience of hand to point out assembling, because of The assembling accuracy rate of this mechanical hand is assembled not as manpower on the contrary, and production efficiency cannot be improved.

Utility model content

In order to overcome shortcoming that prior art exists with not enough, this utility model provides a kind of based on many camera lenses Puma manipulator.

This utility model adopts the following technical scheme that

A kind of puma manipulator based on many camera lenses, including multi-joint multifunction manipulator, to be installed for gathering Joining the ccd video camera of pcb board image, biological contact making device and computer, described ccd video camera is arranged on On hand, biological contact making device is arranged on the finger tip of multi-joint multifunction manipulator to multi-joint multi-functional mechanical, Described ccd video camera, biological contact making device and multi-joint multifunction manipulator are connected with computer.

Described ccd video camera is specially two, is separately mounted to the left side of multi-joint multifunction manipulator forearm And right side.

Described biological contact making device includes the paster made with sensitive material and for measuring paster deformation and exporting The circuitry for signal measurement of the signal of telecommunication, described paster covers on the finger tip of mechanical hand, described paster and signal Measuring circuit connects, and circuitry for signal measurement is connected with computer, and described circuitry for signal measurement is built in multi-joint Inside multifunction manipulator.

Described paster be shaped as finger cot type.

Described sensitive material is specially conductive rubber.

Model and the parameter of said two ccd video camera are identical, and the coordinate system of two ccd video cameras is altogether Face, and each coordinate axes is placed in parallel.

The positioning assembly method of a kind of puma manipulator, comprises the steps:

S1 pcb board to be assembled moves to mechanical hand front, opens two ccd video cameras acquisitions two to be assembled The image of pcb board, the most left and right image；

The PCB template pre-entered in two pcb board images and computer that S2 computer will obtain is carried out Join, determine the target area that shaped piece is to be assembled in two images；

S3 utilizes binocular location algorithm measurement to obtain the distance of target area to be assembled and mechanical hand finger tip, meter Calculation machine controls mechanical hand and moves to target area to be assembled, completes Primary Location；

Target area after S4 utilizes the biological contact making device of mechanical hand finger tip and Primary Location in S3 connects Touching, by circuitry for signal measurement, the deformation signal of telecommunication that paster produces is exported computer, computer carries out position The fine setting put, it is achieved be accurately positioned, completes assembling.

Described S2 is particularly as follows: be filtered the image of the pcb board to be assembled that two ccd video cameras obtain Making an uproar, recycling edge detection operator extracts the marginal portion of pcb board, removes background parts, then uses and divide Section linear transformation, to image enhaucament, improves the contrast of image, finally uses template matching based on gray value Method determines target area to be assembled.

Described S3 utilize binocular location algorithm measurement obtain target area to be assembled and mechanical hand finger tip away from From, computer controlled machine tool hands moves to target area to be assembled, completes Primary Location, particularly as follows:

S3.1 uses Zhang Zhengyou calibration algorithm to solve inner parameter matrix and the external parameter square of ccd video camera Battle array, then carry out binocular solid calibration, determine the relative position relation between two video cameras, described phase para-position The relation of putting includes spin matrix R and translation vector T；

S3.2 uses gray scale Cross Correlation Matching method based on template matching to complete the pixel of two images Join；

S3.3 utilize binocular location algorithm i.e. binocular range-measurement system image-forming principle record target area to be assembled with Distance l of mechanical hand finger tip；

$\begin{array}{cc}l=\frac{Bf}{x}& x=x_{left}-x_{right}\end{array}$

Wherein, the distance of the projection centre of left and right two video camera is baseline distance B, and impact point A is through by two optical axises During the binocular range-measurement system of parallel left and right cameras composition, image in respectively the A1 point in left CCD image planes and A2 point in right CCD image planes, its position in the image planes of left and right is respectively x_leftAnd x_right, two video cameras Focal length is that to be impact point A be imaged on imaging point in the CCD image planes of left and right respectively by binocular camera for f, x Alternate position spike.

Described S4 is particularly as follows: the biological contact making device of mechanical hand finger tip is after completing Primary Location, with to be assembled Pcb board contacts, and sensitive material can deform upon the change causing being under pressure, and the signal of telecommunication of output changes, meter The position of mechanical hand is adjusted by calculation machine according to the signal of output and the mathematical model of pressure and position, from And reach to be accurately positioned the purpose of assembling shaped piece.

The beneficial effects of the utility model:

(1) this utility model can complete oneself of region to be assembled to pcb board when prosthetic intervention Move and be accurately positioned, complete the assembly work of shaped piece；

(2) image that this utility model uses two CCD camera to gather a width pcb board respectively processes, Meet rapidity and the requirement of real-time of industrial electronic assembling；

(3) this utility model adds biological haptic device in the robot, assembling target area can be carried out essence Determine position, improve the accuracy rate of shaped piece assembling；

(4) this utility model employs the binocular location in digital image processing techniques in the location of mechanical hand Algorithm, can the effectively measuring distance obtained between mechanical hand and target assembly area；

(5) this utility model utilizes the method for template matching to determine the target area that shaped piece assembles, and improves The accuracy of follow-up location work.

Accompanying drawing explanation

Fig. 1 is the structural representation of puma manipulator of the present utility model；

Fig. 2 is the structural representation of this utility model biology contact making device；

Fig. 3 a is the searched figure of template matching method of the present utility model；

Fig. 3 b is matching template schematic diagram of the present utility model；

Fig. 4 is workflow diagram of the present utility model.

Detailed description of the invention

Below in conjunction with embodiment and accompanying drawing, this utility model is described in further detail, but this practicality is new The embodiment of type is not limited to this.

Embodiment

As it is shown in figure 1, a kind of puma manipulator based on many camera lenses, including a multi-joint multi-functional mechanical Hands 1, for gathering the ccd video camera 2 of pcb board image to be assembled, biological contact making device 3 and computer, Described ccd video camera is specially two, is separately mounted to left side and the right side of mechanical hand forearm, video camera Particular location can be arranged on the different parts of mechanical hand according to practical situation, and biological contact making device can will be treated The positional information of assembly area is converted into the signal of telecommunication and feeds back to computer.The number of biological contact can be according to reality Border demand is adjusted.Computer is responsible for the signal of telecommunication of the image gathered and feedback is processed and sent Control signal, commander's mechanical hand completes final location and the work of assembling shaped piece, and described mechanical hand is built-in The control circuit of mechanical hand, control circuit is connected with computer.

Two ccd video camera models, parameter that this utility model uses keep consistent, and basic holding optical axis is put down Row, binocular camera coordinate system are coplanar and each coordinate axes is placed in parallel, the left images size of synchronous acquisition, Ratio is consistent, and the half-tone information of image keeps ratio more complete,

As in figure 2 it is shown, described biological contact making device includes the paster 5 made with sensitive material and for measuring Fingerstall deformation also exports the circuitry for signal measurement 4 of the signal of telecommunication, and described paster 5 covers on the finger tip of mechanical hand, Described paster is connected with circuitry for signal measurement, and circuitry for signal measurement 4 is connected with computer, described signal measurement Circuit is built in inside multi-joint multifunction manipulator.In the present embodiment, paster selects finger cot type, and uses Three pasters, are enclosed within the finger tip of mechanical hand.

Described sensitive material has pressure-electric output characteristic, and pressure experienced can be converted into signal of telecommunication output. In the present embodiment paster be shaped as finger cot type, the sensitive material of employing is conductive rubber, uses this sensitive material The biological contact making device made is typically mounted on the finger tip of mechanical hand as fingerstall, when finger tip contacts is to pcb board Time, changing due to deformation pressure experienced of the sensitive material on biological contact making device, thus cause The change of the output signal of telecommunication.By circuitry for signal measurement, can measure and obtain biological contact making device due to deformation And the output signal of telecommunication changed is sent to computer as output signal, computer is according to being previously obtained Mathematical model i.e. can get current mechanical hand location, and export corresponding control signal and be adjusted.

Fig. 3 a is searched figure, and Fig. 3 b is matching template, puts down on searched figure if matching template overlays Moving, the searched figure of that block under template covers is subgraph, compares the content in subgraph and matching template, if two Person's similarity measure maximum then represents that both contents are consistent, and subgraph now is then for matching area to be found.

As shown in Figure 4, use the positioning assembly method that this mechanical hand realizes, comprise the steps:

S1 obtains the image of pcb board to be assembled.

Pcb board to be assembled is transported to the front of mechanical hand through streamline, regulates the brightness of light source, passes through Two, left and right ccd video camera shoots respectively and obtains marking pcb board image, and the ccd video camera being arranged on left side becomes For left ccd video camera, image is referred to as left image, and the image of right ccd video camera shooting is referred to as right image.

S2 template matching finds target area to be assembled.

There is more noise in the pcb board image collected from industry spot, therefore first has to carry out image Filtering and noise reduction.The image of camera acquisition includes pcb board to be assembled, referred to as prospect, simultaneously on image Further comprises background parts, target area will be carried out identification and analysis, first it will be extracted from background Out, therefore image is carried out image dividing processing.Owing to target (pcb board) is the geometry of rule, Edge detection operator can be utilized to extract the marginal portion of pcb board, go to take background parts.

In order to improve the identification of image, carry out highlighting and strengthening by detailed information and the edge of image, have It is beneficial to the template matching of follow-up target area to be assembled, needs pcb board image is carried out image enhaucament, improve The contrast of image, changes former gray value interval scope according to certain mapping relations, realizes the back of the body whereby The effect that scape image strengthens with target image contrast.This utility model uses piecewise linear transform to be strengthened Method.If the gamma function of original image is that (r, c), tonal range is [0, M to f_f], the image function after conversion (r, c), tonal range is [0, M to be expressed as g_g], transformation for mula can be expressed as:

$\left\{\begin{array}{cc}\frac{M_g-d}{M_f-b}\&lsqb;f(r,c)-b\&rsqb;+d& b\&le;f(r,c)\&le;M_f\\ \frac{d-c}{b-a}\&lsqb;f(r,c)-a\&rsqb;+c& a\&le;f(r,c)<b\\ \frac{c}{a}f(r,c)& 0\&le;f(r,c)\&le;a\end{array}\right.$

Template matching is by calculation template image and the similarity measure of image to be searched, thus at image to be searched In find the process of template image.The process of template matching can be expressed as: the most according to pixels calculation template figure As with the similarity measure of image to be searched, then find the similarity measure region of maximum as matched position, its Principle is as shown in Fig. 3 a and Fig. 3 b.

After pcb board image is carried out image enhaucament, owing to the grey value profile in each region of pcb board image is The most fixing, therefore this utility model have employed template matching method based on gray value.Based on gray value Template matching method using the gray value of entire image as similarity measure, utilize the search strategy that defines according to Order from top to bottom, from left to right searches for qualified region in image to be searched, by setting one The search window of individual appointment size, scans for comparing in the search window.

In image to be searched, the position of object can be described by translation.Template by image t (r, c) represents, Area-of-interest therein is appointed as T, and template matching is exactly according to a definite sequence translation mould in image to be matched Plate area-of-interest T, then calculates this region and the similarity value of template area-of-interest in image to be matched s.Similarity measure is described by following formula:

S (r, c)=s{t (u, v), f (r+u, c+v)；(u,v)∈T}

Wherein s (r, c) represents the similarity measure calculated based on gray value, t (u, v) represents the gray value of each point in template, F (r+u, c+v) represents that template area-of-interest moves on to the gray value of image current location.

The straightforward procedure asking for similarity measure is to calculate the absolute value sum of gray value difference between two images (SAD) or the quadratic sum (SSD) of all differences, SAD and SSD can represent by following two formulas respectively:

$sad(r,c)=\frac{1}{n}\underset{(u,v)\&Element;T}{\&Sigma;}|t(u,v)-f(r+u,c+v)|$

$ssd(r,c)=\frac{1}{n}\underset{(u,v)\&Element;T}{\&Sigma;}{\&lsqb;t(u,v)-f(r+u,c+v)\&rsqb;}^2$

Wherein, n represents the quantity of pixel, i.e. n=| T | in this interest region of template.SAD and SSD is come Saying, the value of similarity measure is the biggest, and the difference between image to be searched and template is the biggest.Use based on ash The template matching method of angle value i.e. can determine that target area to be assembled.

S3 utilizes binocular location algorithm measurement to obtain the distance of target area to be assembled and mechanical hand finger tip, meter Calculation machine controls mechanical hand and moves to target area to be assembled, completes Primary Location；

Binocular camera is demarcated by S3.1.

Owing to using parameter identical two video cameras composition binocular solid system herein, therefore carry Go out and the most respectively two video cameras have been calibrated, solved its inside and outside parameter.Image coordinate system and world coordinates Transformational relation between system is:

$\begin{array}{c}s\left[\begin{array}{c}u\\ v\\ 1\end{array}\right]=\left[\begin{array}{cccc}{f}_x& 0& u_0& 0\\ 0& f_y& v_0& 0\\ 0& 0& 1& 0\end{array}\right]\left[\begin{array}{cc}{R}_{3\&times;3}& T_{3\&times;1}\\ 0^T& 1\end{array}\right]\left[\begin{array}{c}{X}_w\\ Y_w\\ Z_w\\ 1\end{array}\right]\\ =AM\left[\begin{array}{c}{X}_w\\ Y_w\\ Z_w\\ 1\end{array}\right]=H\left[\begin{array}{c}{X}_w\\ Y_w\\ Z_w\\ 1\end{array}\right]\end{array}$

A is the Intrinsic Matrix of video camera, and M is determined relative to position and the direction of world coordinate system by video camera Fixed, unrelated with the inner parameter figure pinhole camera modeling of video camera, M is called video camera external parameter matrix, Zhang Zhengyou calibration algorithm is used to solve inner parameter matrix and the external parameter matrix of video camera.

Completing the demarcation to single camera, after obtaining the inside and outside parameter of single camera, then it is vertical to carry out a binocular Body is calibrated, and about calculating, the relative position between two video cameras is outer parameter, and i.e. left and right two video cameras is relative Position relationship, including spin matrix R and translation vector T.

Assume that the three-dimensional world coordinate putting P on gridiron pattern scaling board is X_w, use two video cameras collection figure simultaneously Picture, P point coordinate under left and right cameras coordinate system is respectively X_LAnd X_R, the outer ginseng matrix of left and right cameras It is respectively (R_L,t_L) and (R_R,t_R), have according to the transformational relation of world coordinate system with camera coordinate system:

$\left\{\begin{array}{c}{X}_L=R_L{X}_W+t_L\\ X_R=R_R{X}_W+t_R\end{array}\right.$

Can obtain from left video camera to the transformation relation of right video camera according to above formula:

X_R=RX_L+T

Wherein

R=R_RR_L ^-1, T=t_R-Rt_L

So just can obtain binocular vision system neutral body camera according to the respective outer parameter of left and right cameras The position relationship of Relation Parameters, i.e. stereoscopic camera, so far completes the staking-out work of binocular camera.

S3.2 uses gray scale Cross Correlation Matching method based on template matching to complete the pixel coupling of two images；

Left and right cameras model that this utility model is used, parameter keep consistent, basic keep optical axis parallel, Binocular camera coordinate system is coplanar and each coordinate axes is placed in parallel, the left images size of synchronous acquisition, ratio Unanimously, the target area image that the half-tone information of image keeps ratio more complete and to be assembled determines in S2. Therefore this utility model uses gray scale Cross Correlation Matching method based on template matching to complete object region High accuracy coupling.

Normalized crosscorrelation matching algorithm is according to searching for foundation between subgraph on template image and image to be matched Cross-correlation function judges whether coupling, uses cross-correlation function expression formula as follows:

$N(i,j)=\frac{\underset{m=1}{\overset{M}{\&Sigma;}}\underset{n=1}{\overset{N}{\&Sigma;}}T(m,n)S^{i,j}(m,n)}{\sqrt{\underset{m=1}{\overset{M}{\&Sigma;}}\underset{n=1}{\overset{N}{\&Sigma;}}{T}^2(m,n)\underset{m=1}{\overset{M}{\&Sigma;}}\underset{n=1}{\overset{N}{\&Sigma;}}{\&lsqb;S^{i,j}(m,n)\&rsqb;}^2}}$

$N(i,j)=\frac{\underset{m=1}{\overset{M}{\&Sigma;}}\underset{n=1}{\overset{N}{\&Sigma;}}\left(T\right(m,n)-\stackrel{\&OverBar;}{T(m,n)})\left(S^{i,j}\right(m,n)-\stackrel{\&OverBar;}{S^{i,j}(m,n)})}{{\sqrt{\underset{m=1}{\overset{M}{\&Sigma;}}\underset{n=1}{\overset{N}{\&Sigma;}}{\&lsqb;T(m,n)-\stackrel{\&OverBar;}{T(m,n)}\&rsqb;}^2\underset{m=1}{\overset{M}{\&Sigma;}}\underset{n=1}{\overset{N}{\&Sigma;}}\&lsqb;S^{i,j}(m,n)-\stackrel{\&OverBar;}{S^{i,j}(m,n)}\&rsqb;}}^2}$

$\stackrel{\&OverBar;}{T(m,n)}=\frac{1}{M\&times;N}\underset{m=1}{\overset{M}{\&Sigma;}}\underset{n=1}{\overset{N}{\&Sigma;}}T(m,n)$

$\stackrel{\&OverBar;}{S^{i,j}(m,n)}=\frac{1}{M\&times;N}\underset{m=1}{\overset{M}{\&Sigma;}}\underset{n=1}{\overset{N}{\&Sigma;}}{S}^{i,j}(m,n)$

In above formula, template image be T (m, n), template image size is M × N,For T (m, n) on institute There is the meansigma methods of pixel grey scale；On a reference so that (i is j) that the search image-region of top left corner pixel point is S^i,j(m, n),Meansigma methods for pixel grey scales all on search graph picture.Images match is coupling target figure Top left corner pixel point as the i.e. template in region.Cross-correlation function value N (i, span j) Be 0≤N (i, j)≤1, the size of its value depends on reference picture so that (i j) is the search of top left corner pixel point Image-region and the matching degree of template image.The cross-correlation function value that certain pixel is corresponding is the biggest, and this is described Pixel matching degree is the highest, chooses the pixel that cross-correlation function value is maximum, is matched pixel point.

After the coupling completing left images pixel, as it can be seen, former according to binocular range-measurement system imaging Reason, the ccd video camera that left and right two model is consistent is placed in parallel at grade, the throwing of left and right two video camera The distance at shadow center is baseline distance B.

Impact point A, through the binocular range-measurement system being made up of the left and right cameras that two optical axises are parallel, becomes respectively As the A1 point in left CCD image planes and the A2 point in right CCD image planes, its position in the image planes of left and right is divided Wei x_leftAnd x_right.Known two focal length of camera are f, can derive tested according to Similar Principle of Triangle Distance l:

$\begin{array}{cc}l=\frac{Bf}{x}& x=x_{left}-x_{right}\end{array}$

X is A point is imaged on the alternate position spike of imaging point in the CCD image planes of left and right respectively by binocular camera, again by It is referred to as binocular parallax.Visible at binocular camera optical axis perfect parallelism ideally, and obtain simultaneously The image of target object, determines same target corresponding positions in the ccd image of left and right by image matching algorithm Putting, calculate binocular parallax x, oneself knows focal length and baseline size again, can obtain target range by above formula, So far the first location of target area to be assembled is completed.

Target area after S4 utilizes the biological contact making device of mechanical hand finger tip and Primary Location in S3 connects Touching, by circuitry for signal measurement, the deformation signal of telecommunication that paster produces is exported computer, computer carries out position The fine setting put, it is achieved be accurately positioned, completes assembling.

Mechanical hand, when assembling, can contact with pcb board, and the finger tip of mechanical hand can be under pressure Effect.When abnormity components and parts accurately contact with region to be assembled and special-shaped components and parts touch its of pcb board During its region, pressure experienced effect is different, therefore can set up the mathematical model of pressure and position, logical Cross finger tip pressure experienced effect to adjust the position of mechanical hand, thus complete the essence to target area to be assembled Determine position.

This utility model have employed a kind of sensitive material and devises biological contact making device, and this sensitive material has pressure -electric output characteristic, can be converted into signal of telecommunication output by pressure experienced.The biology made with this sensitive material Contact making device is typically mounted on the finger tip of mechanical hand as fingerstall, when finger tip contacts to pcb board, biological tactile Changing due to deformation pressure experienced of conductive material on some device, thus cause exporting the signal of telecommunication Change.Between sensitive material and mechanical hand finger tip, there is circuitry for signal measurement, can measure and obtain biological touching The output signal of telecommunication that some device changes due to deformation is sent to computer as output signal, computer I.e. can get current mechanical hand location according to the mathematical model being previously obtained, and export corresponding control Signal is adjusted.

Mechanical hand completes the first location in region to be assembled in S3 step, mobile manipulator to this region, Mechanical hand and pcb board come in contact after according to the output signal of the biological contact making device on finger tip to mechanical hand Position is finely adjusted, and is finally reached the assembling being accurately positioned and completing shaped piece.

Above-described embodiment is this utility model preferably embodiment, but embodiment of the present utility model is not subject to The restriction of described embodiment, other any without departing from being made under spirit of the present utility model and principle Change, modify, substitute, combine, simplify, all should be the substitute mode of equivalence, be included in this practicality new Within the protection domain of type.

Claims (6)

1. a puma manipulator based on many camera lenses, it is characterized in that, including multi-joint multifunction manipulator, for gathering the ccd video camera of pcb board image to be assembled, biological contact making device and computer, described ccd video camera is arranged on multi-joint multi-functional mechanical on hand, biological contact making device is arranged on the finger tip of multi-joint multifunction manipulator, and described ccd video camera, biological contact making device and multi-joint multifunction manipulator are connected with computer.

Puma manipulator the most according to claim 1, it is characterised in that described ccd video camera is specially two, is separately mounted to left side and the right side of multi-joint multifunction manipulator forearm.

Puma manipulator the most according to claim 1, it is characterized in that, described biological contact making device includes the paster made with sensitive material and for measuring paster deformation and exporting the circuitry for signal measurement of the signal of telecommunication, described paster covers on the finger tip of mechanical hand, described paster is connected with circuitry for signal measurement, circuitry for signal measurement is connected with computer, and described circuitry for signal measurement is built in inside multi-joint multifunction manipulator.

Puma manipulator the most according to claim 3, it is characterised in that described paster be shaped as finger cot type.

Puma manipulator the most according to claim 3, it is characterised in that described sensitive material is specially conductive rubber.

Puma manipulator the most according to claim 2, it is characterised in that model and the parameter of said two ccd video camera are identical, and the coordinate system of two ccd video cameras is coplanar, and each coordinate axes is placed in parallel.