CN103837085A - Target displacement vector measuring device and method based on laser tracker point-by-point calibration - Google Patents

Abstract

The invention provides a target displacement vector measuring device and method based on laser tracker point-by-point calibration, and belongs to the technical field of infinitesimal displacement measurement. The target displacement vector measuring device and method aim to solve the problem that measurement results obtained through a method for detecting target infinitesimal displacement through an existing visual-inspection measuring system are low in accuracy. The device comprises a high-speed CCD video camera, a laser tracker and a diffuse reflection projection screen, wherein a target object moves on the diffuse reflection projection screen. According to the method, first, a corresponding relation look-up table is established based on the principle that imaging points of target object calibration points and target object calibration points acquired through the laser tracker are conjugated; then, the high-speed CCD video camera is used for sequentially imaging screen light spots of the target object, a conjugate point of a target object projection point is acquired, and the coordinate value of the target object projection point on a projection screen coordinate system is obtained through table look-up or a resolving method combining table look-up with bilinear interpolation. The target displacement vector measuring device and method based on laser tracker point-by-point calibration are used for measuring a target displacement vector.

Description

Displacement of targets device for measuring vector quantity and the method for demarcating based on laser tracker pointwise

Technical field

The present invention relates to displacement of targets device for measuring vector quantity and the method for demarcating based on laser tracker pointwise, belong to micro-displacement detection technique field.

Background technology

Micro-displacement detection technique is significant in modern precision fields of measurement, by the high-acruracy survey to micro-displacement, and in conjunction with the structural parameters of space layout, can complete the precision measurement to a series of important parameters such as angle, speed, linearity, flatnesses.Simultaneously it can also be for assessment of the control accuracy of all kinds of existing instruments, and can detect the mobile accuracy of objective body in space.

The displacement measuring technology that precision is higher in the world is at present double-frequency laser interferometry method, and its displacement accuracy recording is nanoscale.But this method can only be measured the one dimension scalar shift value of prism of corner cube, cannot measure two-dimensional vector shift value, and can not follow the tracks of concrete target in kind, and visual field is very little, has in the use larger limitation.

At present in fields of measurement, conventionally adopt vision detection technology to measure the positional information of space object, this measuring method cost is low, precision is high, visual field is large, and be convenient to installation and transportation, strong adaptability, then coordinate high-precision image processing algorithm can realize high speed to objective body, and high-precision location and displacement measuring and calculating on a large scale.

In typical vision-based detection measurement scheme, if adopting the optical imagery equation of perspective model measures the displacement of the lines vector of object space, this measuring system is generally made up of light source, camera lens, ccd video camera, image collecting device and image processing software, and it calculates the space position parameter of object by visual imaging theory.Measuring principle as shown in Figure 2.

The principle of typical vision measurement displacement is: ccd video camera utilizes camera lens to carry out imaging to the object point A on object plane screen, forms picture point B in CCD image planes.When target object point on screen from A ₁point moves to A ₂point, imaging point in CCD image planes from B ₁point moves to B ₂point.Can calculate picture point from B by follow-up image processing algorithm ₁point moves to B ₂point displacement be x, target object point on screen from A ₁point moves to A ₂the displacement of point is d, and the satisfied pass of each parameter is:

$d=x\frac{l}{f}.$

Wherein, the photocentre that l is imaging system is to the vertical range of screen 4, and f is the lens focus of ccd video camera 2.Method can calculate the displacement vector value of objective body on screen 4 easily accordingly.There is following defect in the preparation method of this displacement vector value:

1, manufacturability defect: CCD camera and camera lens, in manufacture process, exist manufacturability defect, there is certain error with the setting value f dispatching from the factory in its lens focus, and this error is introduced and is measured equation, can cause measuring error.And the accurate location of imaging system photocentre is difficult to definite, and then is difficult to provide the accurate distance l of photocentre to screen in actual measurement, and this all causes error effect to measurement result.

If 2 perspective model errors: if perspective model is fairly simple, there is no the problem such as considering screen and primary optical axis angle, when screen plane out-of-flatness, or screen and primary optical axis be while existing angle, and the imaging formula of system is all untenable.And in actual conditions, there is distortion in camera lens, can cause very large nonlinearity erron, the kind difference of distortion, the error causing distributes also different, and the way being difficult to by demarcating is corrected problems, has limited thus detectability and the accuracy of detection of vision detection system to target micrometric displacement.

Summary of the invention

The present invention seeks in order to solve the detection method of existing vision-based detection measuring system to target micrometric displacement, the low problem of measurement result precision of existence, provides a kind of displacement of targets device for measuring vector quantity and method of demarcating based on laser tracker pointwise.

The displacement of targets device for measuring vector quantity of demarcating based on laser tracker pointwise of the present invention, it comprises high speed CCD camera, laser tracker and diffuse reflection projection screen, object moves on diffuse reflection projection screen;

High speed CCD camera is for the object calibration point on diffuse reflection projection screen is carried out to imaging, and laser tracker is for measuring the object calibration point coordinate on diffuse reflection projection screen.

The square of described diffuse reflection projection screen for being spliced by the ground glass of two 1000mm × 500mm × 3mm, the back spraying coating black paint vehicle of this diffuse reflection projection screen.

Based on the measuring method of the above-mentioned displacement of targets device for measuring vector quantity of demarcating based on laser tracker pointwise, it comprises the following steps:

Step 1: on setting diffuse reflection projection screen, 1 A1 is as object calibration point, and obtaining the coordinate figure of this object calibration point A1 on projection screen coordinate system by laser tracker is A1 (U1, V1);

Step 2: keep object calibration point A1 to fix, utilize high speed CCD camera to carry out imaging to this object calibration point A1, in the image planes of high speed CCD camera, obtain the corresponding conjugate points B1 of object calibration point A1, conjugate points B1 is carried out after image processing, obtain the coordinate figure B1 (m1, n1) that conjugate points B1 fastens at image coordinates;

Step 3: on diffuse reflection projection screen, taking object calibration point A1 as benchmark, more selected 24 object calibration point Ai, i=2,3,4 ... 25,25 object calibration points spacing with 25cm on diffuse reflection projection screen is evenly arranged, then successively according to the method for step 1 and step 2, until complete the measuring and calculating to the 25th object calibration point, obtain Ai (Ui, Vi) and Bi (mi, ni) corresponding relation look-up table, i=1,2,3 ..., 25;

Step 4: adopt high speed CCD camera successively to the object screen hot spot imaging that projection obtains on diffuse reflection projection screen, obtain the conjugate points Bj (mj, nj) of object subpoint, j=1,2,3,

Step 5: search conjugate points Bj (mj, nj) in described corresponding relation look-up table, if Bj (mj, nj) overlaps with certain the some Bi (mi, ni) in corresponding relation look-up table, perform step six; Otherwise execution step seven;

Step 6: obtain the coordinate figure of object subpoint on projection screen coordinate system according to corresponding relation look-up table, then perform step eight;

Step 7: in described corresponding relation look-up table according to Bj (mj, nj) immediate four known calibration point Bi (mi around, ni), adopt the calculation method of bilinear interpolation to obtain Aj (Uj, Vj), Aj (Uj, Vj) is the object subpoint corresponding with conjugate points Bj (mj, the nj) coordinate figure on projection screen coordinate system; Then perform step eight;

Step 8: initial value and final value by step 4 to the Aj (Uj, Vj) obtaining in step 7 process, calculate and obtain the displacement vector of object on diffuse reflection projection screen.

Described object calibration point is hot spot or target ball.

Described conjugate points B1 is carried out after image processing, obtain the coordinate figure B1 (m1 that conjugate points B1 fastens at image coordinates, n1) concrete grammar is: the 8 bit image gradation datas of the conjugate points B1 first high speed CCD camera being obtained carry out dual threshold pixel extraction, in the effective pixel area of conjugate points B1, utilize the method for linear interpolation to increase valid pixel number, the barycenter extraction algorithm of recycling based on intensity-weighted calculates and obtains the coordinate figure B1 (m1, n1) that conjugate points B1 fastens at image coordinates.

In the effective pixel area of conjugate points B1, utilize the method for linear interpolation to increase valid pixel number, the detailed process of the coordinate figure B1 (m1, n1) that the barycenter extraction algorithm calculating acquisition conjugate points B1 of recycling based on intensity-weighted fastens at image coordinates is:

Below taking hypothesis a bit as example, this detailed process is described:

Pre-set the gradation of image database of conjugate points, this database comprises the multiple pixel f (p that are matrix arrangement, q), p is the line number of matrix, q is matrix column number, the 8 bit image gradation datas of the conjugate points B1 that setting high-speed ccd video camera obtains are f (u, v), and these gradation of image data f (u, v) are positioned at pixel f (2,2), the f (2 of gradation of image database, 3), f (3,2) and between f (3,3)

Utilize the method acquisition f (u, 2) of linear interpolation to be:

f(u,2)=f(2,2)+(u-2)·[f(3,2)-f(2,2)]，

Obtaining f (u, 3) by f (2,3) and f (3,3) interpolation is again:

f(u,3)=f(2,3)+(u-2)·[f(3,3)-f(2,3)]，

Finally, utilize f (u, 2) and f (u, 3) to carry out interpolation, obtain f (u, v):

f(u,v)=f(u,2)+(v-2)·[f(u,3)-f(u,2)]；

The pixel f (2 of former formation hot spot, 2), f (2,3), f (3,2) and f (3,3) are by after interpolation, become f (2,2), f (2,3), f (3,2), f (3,3) and f (u, v);

Based on this, the coordinate figure B1 (m1, n1) that utilizes the barycenter extraction algorithm calculating acquisition conjugate points B1 based on intensity-weighted to fasten at image coordinates is:

$m1=\frac{2\·f\left(\mathrm{2,2}\right)+2\·f\left(\mathrm{2,3}\right)+3\·f\left(\mathrm{3,2}\right)+3\·f\left(\mathrm{3,3}\right)+u\·f(u,v)}{f\left(\mathrm{2,2}\right)+f\left(\mathrm{2,3}\right)+f\left(\mathrm{3,2}\right)+f\left(\mathrm{3,3}\right)+f(u,v)},$

$\mathrm{<figure-callout\; id="1"\; label="n"\; filenames="HDA0000474152030000011.png"\; state="\{\{state\}\}">n</figure-callout>}1=\frac{2\&CenterDot;f\left(\mathrm{2,2}\right)+3\&CenterDot;f\left(\mathrm{2,3}\right)+2\&CenterDot;f\left(\mathrm{3,2}\right)+3\&CenterDot;f\left(\mathrm{3,3}\right)+v\&CenterDot;f(u,v)}{f\left(\mathrm{2,2}\right)+f\left(\mathrm{2,3}\right)+f\left(\mathrm{3,2}\right)+f\left(\mathrm{3,3}\right)+f(u,v)}.$

Advantage of the present invention: a kind of vision detection technology that the invention belongs to micro-displacement measurement technical field, its demarcation based on laser tracker is calculated, the advantage that it had both kept the large visual field of vision detection technology dynamically to follow the tracks of at a high speed, has also effectively promoted the accuracy of detection to micro-displacement.

The present invention, taking vision detection technology as basis, has solved the multinomial problem in vision detection technology cleverly, and the relation of searching with correspondence and linear difference algorithm have substituted original optical measurement equation, have directly eliminated polynomial system error.And the bilinearity difference arithmetic adopting, has effectively eliminated all linear errors in system, is a kind of novel Micro-Displacement.

Apparatus of the present invention and method have utilized laser tracker to measure application to the pinpoint characteristic of objective body cleverly, have improved the service efficiency of laser tracker.

The present invention has effectively solved the test problems of the displacement of targets vector under large visual field environment, has met the testing requirement of high precision and large visual field simultaneously.Reality can reach 0.2mm to the positioning precision of hot spot, and field of detection is 1m, and accuracy of detection is 0.02% relatively.In addition this detection method also has dynamic tracking effect at a high speed, application in practice.

Brief description of the drawings

Fig. 1 is the principle schematic of the displacement of targets device for measuring vector quantity of demarcating based on laser tracker pointwise of the present invention;

Fig. 2 is the schematic diagram that typical vision measuring system is measured displacement;

Fig. 3 is the schematic diagram that increases valid pixel by the method for linear interpolation;

Fig. 4 adopts the method for bilinear interpolation to table look-up to obtain the schematic diagram of Aj (Uj, Vj) in step 7.

Embodiment

Embodiment one: present embodiment is described below in conjunction with Fig. 1, the displacement of targets device for measuring vector quantity of demarcating based on laser tracker pointwise described in present embodiment, it comprises high speed CCD camera 1, laser tracker 2 and diffuse reflection projection screen 3, and object moves on diffuse reflection projection screen 3;

High speed CCD camera 1 carries out imaging for the object calibration point on diffuse reflection projection screen 3, and laser tracker 2 is for measuring the object calibration point coordinate on diffuse reflection projection screen 3.

In present embodiment, the space ranging scope of laser tracker 2 is not less than 50m, and measurable angle range is better than 100 ° in the vertical direction, and horizontal direction is better than 180 °, and positioning precision is better than 20 μ m+5 μ m/m.

The lens focus of high speed CCD camera 1 is 16mm, and single pixel dimension is 7 μ m × 7 μ m, and resolution is 2560 pixel × 1920 pixels, and image planes are of a size of 16.0mm × 12.1mm, and frame frequency 1130 frames are per second.

The principle of work of measurement mechanism is: within the scope of apparent field, high speed CCD camera 1 and diffuse reflection projection screen 3 form a set of optical imaging system.Therefore, any object point A on diffuse reflection projection screen 3 after optical imagery, can form corresponding picture point B in the image planes of high speed CCD camera 1.And A point and B point conjugation,, there is conjugate relation one to one in the each picture point in each object point and CCD image planes on diffuse reflection projection screen 3.There is unique corresponding relation in the image coordinates system that is projection screen coordinate system and high speed CCD camera 1.

Embodiment two: present embodiment is described further embodiment one, the square of diffuse reflection projection screen 3 for being spliced by the ground glass of two 1000mm × 500mm × 3mm described in present embodiment, the back spraying coating black paint vehicle of this diffuse reflection projection screen 3.

In present embodiment, the surface area of diffuse reflection projection screen 3 is 1m × 1m, the black paint vehicle at its back side is used for eliminating secondary reflecting.The flatness of ground glass requires to be better than 0.2mm, and between two blocks of ground glass, seam is less than 0.5mm, and the flatness of diffuse reflection projection screen 3 entirety is better than 0.5mm, and ground verticality is better than 0.1 °.

Embodiment three: below in conjunction with Fig. 1, present embodiment is described, the measuring method based on the above-mentioned displacement of targets device for measuring vector quantity of demarcating based on laser tracker pointwise described in present embodiment, it comprises the following steps:

Step 1: on setting diffuse reflection projection screen 3,1 A1 is as object calibration point, and obtaining the coordinate figure of this object calibration point A1 on projection screen coordinate system 4 by laser tracker 2 is A1 (U1, V1);

Step 2: keep object calibration point A1 to fix, utilize high speed CCD camera 1 to carry out imaging to this object calibration point A1, in the image planes of high speed CCD camera 1, obtain the corresponding conjugate points B1 of object calibration point A1, conjugate points B1 is carried out after image processing, obtaining conjugate points B1 is the coordinate figure B1 (m1, n1) on 5 at image coordinates;

Step 3: on diffuse reflection projection screen 3, taking object calibration point A1 as benchmark, more selected 24 object calibration point Ai, i=2,3,4 ... 25,25 object calibration points spacing with 25cm on diffuse reflection projection screen 3 is evenly arranged, then successively according to the method for step 1 and step 2, until complete the measuring and calculating to the 25th object calibration point, obtain Ai (Ui, Vi) and Bi (mi, ni) corresponding relation look-up table, i=1,2,3 ..., 25;

Step 4: adopt high speed CCD camera 1 successively to the object screen hot spot imaging that projection obtains on diffuse reflection projection screen 3, obtain the conjugate points Bj (mj, nj) of object subpoint, j=1,2,3,

Step 5: search conjugate points Bj (mj, nj) in described corresponding relation look-up table, if Bj (mj, nj) overlaps with certain the some Bi (mi, ni) in corresponding relation look-up table, perform step six; Otherwise execution step seven;

Step 6: obtain the coordinate figure of object subpoint on projection screen coordinate system 4 according to corresponding relation look-up table, then perform step eight;

Step 7: in described corresponding relation look-up table according to Bj (mj, nj) immediate four known calibration point Bi (mi around, ni), adopt the calculation method of bilinear interpolation to obtain Aj (Uj, Vj), Aj (Uj, Vj) is the object subpoint corresponding with conjugate points Bj (mj, the nj) coordinate figure on projection screen coordinate system 4; Then perform step eight;

Step 8: initial value and final value by step 4 to the Aj (Uj, Vj) obtaining in step 7 process, calculate and obtain the displacement vector of object on diffuse reflection projection screen 3.

In present embodiment, complete conjugate pair and answer after look-up table, can measure.In measurement, high speed CCD camera 1, to the imaging of screen hot spot, obtains the conjugate points of object subpoint, then in corresponding relation look-up table, searches.If conjugate points is not demarcated in look-up table, adopt the calculation method based on bilinear interpolation, obtain its coordinate figure on projection screen coordinate system 4 according to its four known calibration points around.

Finally, draw the displacement vector of object on diffuse reflection projection screen 3 according to the variation of Aj (Uj, Vj) by simple analytic relationship, realized thus the high-precision two-dimentional miniature linear displacement in large visual field and detected.

The measuring principle basis of present embodiment is the corresponding conjugate relation of object point and picture point in imaging system, and it has set up the corresponding search relationship of object plane coordinate system and image coordinates system, is a kind of taking the theoretical algorithm of tabling look-up for assisting of bilinear interpolation.

Embodiment four: present embodiment is described further embodiment three, object calibration point is hot spot or target ball described in present embodiment.

Embodiment five: present embodiment is described further embodiment four, described in present embodiment, conjugate points B1 is carried out after image processing, obtaining conjugate points B1 is the coordinate figure B1 (m1 on 5 at image coordinates, n1) concrete grammar is: the 8 bit image gradation datas of the conjugate points B1 first high speed CCD camera 1 being obtained carry out dual threshold pixel extraction, in the effective pixel area of conjugate points B1, utilize the method for linear interpolation to increase valid pixel number, the barycenter extraction algorithm of recycling based on intensity-weighted calculates and obtain conjugate points B1 is the coordinate figure B1 (m1 on 5 at image coordinates, n1).

In present embodiment, because the gradation of image value of hot spot region is higher, if the 8 bit image gradation datas of the conjugate points B1 obtaining are arranged as from high to low by gray-scale value: 255, 255, 236, 221, 219, 219, 217, 216, 209, 208, 201, 197, 164, 133, 112 in above-mentioned Grey imaging data: the barycenter region that the pixel that gray-scale value is 255 is hot spot, gray-scale value is that the pixel of 236-197 is hot spot endocyclic area, gray-scale value is that the point of 164-112 is hot spot outer region, it is background area that gray-scale value is less than 100 point, the gray scale interval value in above-mentioned each region is the value of drafting, actual segmentation threshold produces according to on-the-spot illumination debugging.For improving the positioning precision of algorithm to facula mass center, in reality, adopt the pixel data of hot spot endocyclic area to carry out the position extraction of facula mass center.Set gray scale higher limit by algorithm, upper threshold is 236, and gray scale lower limit is limited to 197 under threshold value, selects the pixel of gray-scale value between two threshold values to carry out follow-up facula mass center and extracts, be i.e. Double Thresholding Segmentation method.

The pixel coordinate extracting through Double Thresholding Segmentation method is the approximate region at hot spot place, and in actual imaging, screen distance camera is far away, and the valid pixel of camera after to hot spot imaging is less, and computational accuracy is inadequate.And then the coordinate figure of facula mass center is calculated in employing according to gray scale square weighting centroid algorithm, in hot spot imaging approximate region, utilize linear interpolation to increase valid pixel number and then improve the positioning precision of facula mass center, the coordinate precision that the hot spot providing after barycenter resolves is 5 at image coordinates is better than the physical picture element size of 10 times.

Described image is processed for realizing the precise positioning of screen hot spot and dynamically following the tracks of, pass through image processing algorithm, can give the real time position coordinate information of glossing up on projection screen coordinate system, and can follow the tracks of at a high speed the movement locus of hot spot, provide track fitting image.Realize dynamically following the tracks of and static high-acruracy survey hot spot on a large scale simultaneously.It both can complete the high-acruracy survey to hot spot static position coordinate, can complete again in large visual field the high speed of target body movements track is followed the tracks of, and provided the displacement value that objective body moves simultaneously.

Embodiment six: present embodiment is described below in conjunction with Fig. 1, Fig. 3 and Fig. 4, present embodiment is described further embodiment five, described in present embodiment, in the effective pixel area of conjugate points B1, utilize the method for linear interpolation to increase valid pixel number, the barycenter extraction algorithm of recycling based on intensity-weighted calculates the detailed process that acquisition conjugate points B1 is the coordinate figure B1 (m1, n1) on 5 at image coordinates:

Below taking hypothesis a bit as example, this detailed process is described:

Pre-set the gradation of image database of conjugate points, this database comprises the multiple pixel f (p that are matrix arrangement, q), p is the line number of matrix, q is matrix column number, the 8 bit image gradation datas of the conjugate points B1 that setting high-speed ccd video camera 1 obtains are f (u, v), and these gradation of image data f (u, v) are positioned at pixel f (2,2), the f (2 of gradation of image database, 3), f (3,2) and between f (3,3)

Utilize the method acquisition f (u, 2) of linear interpolation to be:

f(u,2)=f(2,2)+(u-2)·[f(3,2)-f(2,2)]，

Obtaining f (u, 3) by f (2,3) and f (3,3) interpolation is again:

f(u,3)=f(2,3)+(u-2)·[f(3,3)-f(2,3)]，

Finally, utilize f (u, 2) and f (u, 3) to carry out interpolation, obtain f (u, v):

f(u,v)=f(u,2)+(v-2)·[f(u,3)-f(u,2)]；

The pixel f (2 of former formation hot spot, 2), f (2,3), f (3,2) and f (3,3) are by after interpolation, become f (2,2), f (2,3), f (3,2), f (3,3) and f (u, v); This pixel increases mode, can improve the precision of follow-up computing.

Based on this, utilizing barycenter extraction algorithm based on intensity-weighted to calculate to obtain conjugate points B1 is that coordinate figure B1 (m1, n1) on 5 is at image coordinates:

$\mathrm{<figure-callout\; id="1"\; label="m"\; filenames="HDA0000474152030000011.png"\; state="\{\{state\}\}">m</figure-callout>}1=\frac{2\&CenterDot;f\left(\mathrm{2,2}\right)+2\&CenterDot;f\left(\mathrm{2,3}\right)+3\&CenterDot;f\left(\mathrm{3,2}\right)+3\&CenterDot;f\left(\mathrm{3,3}\right)+u\&CenterDot;f(u,v)}{f\left(\mathrm{2,2}\right)+f\left(\mathrm{2,3}\right)+f\left(\mathrm{3,2}\right)+f\left(\mathrm{3,3}\right)+f(u,v)},$

$\mathrm{<figure-callout\; id="1"\; label="n"\; filenames="HDA0000474152030000011.png"\; state="\{\{state\}\}">n</figure-callout>}1=\frac{2\&CenterDot;f\left(\mathrm{2,2}\right)+3\&CenterDot;f\left(\mathrm{2,3}\right)+2\&CenterDot;f\left(\mathrm{3,2}\right)+3\&CenterDot;f\left(\mathrm{3,3}\right)+v\&CenterDot;f(u,v)}{f\left(\mathrm{2,2}\right)+f\left(\mathrm{2,3}\right)+f\left(\mathrm{3,2}\right)+f\left(\mathrm{3,3}\right)+f(u,v)}.$

In the inventive method, if CCD camera imaging system obtains the pixel coordinate Bi (mi, ni) of a certain hot spot, the volume coordinate Ai (Ui, Vi) of hot spot on its unique corresponding screen.It tables look-up principle sketch as shown in Figure 4:

The principle of setting up according to look-up table, there is one-to-one relationship in volume coordinate Ai (Ui, Vi) and the hot spot pixel coordinate Bi (mi, ni) on imaging system image planes of hot spot on screen.

After system building completes, first demarcate, in sketch, 1-4 point is calibration point, has A1 (U1, V1), A2 (U2, V2), A3 (U3, V3), A4 (U4, V4) and B1 (m1, n1), B2 (m2, n2), B3 (m3, n3), B4 (m4, n4) one-to-one relationship, deposits this relation in appointed area, sets up demarcation look-up table.

After demarcation completes, system can be carried out detection and location and follow the tracks of at a high speed the hot spot on screen, in actual detection, imaging system is to the hot spot imaging on screen, the pixel coordinate Bi (mi, ni) that obtains hot spot imaging by algorithm, brings this pixel coordinate into look-up table and searches:

If 1 Bi (mi, ni) is by chance B1 (m1, n1), B2 (m2, n2), B3 (m3, n3), B4 (m4, n4) or known calibration point, can directly in look-up table, obtain the volume coordinate A1 (U1, V1) that hot spot is corresponding, A2 (U2, V2), A3 (U3, V3)

A4(U4,V4)。

If 2 Bi (mi, ni) are not known calibration points, in look-up table, cannot find its corresponding volume coordinate Ai (Ui, Vi).Suppose that Bi (mi, ni) is B5 (m5, n5), cannot directly provide its corresponding volume coordinate A5 (U5, V5) by look-up table.Construct interpolation algorithm, utilize known A1 (U1, V1), A2 (U2, V2), A3 (U3, V3), A4 (U4, V4) with B1 (m1, n1), B2 (m2, n2), B3 (m3, n3), the corresponding relation of B4 (m4, n4), structure interpolating function A5 (U5, V5)=f[B (m5, n5)], calculate the volume coordinate A5 (U5, V5) that B5 (m5, n5) is corresponding.

The inventive method, if coordinate the time interval of twice detection, can calculate the movement velocity of target hot spot; Can also pass through the coordinate figure of algorithm stores hot spot, and then record the running orbit of hot spot.It does not use Optical Formula in testing process, and then has avoided the impact of the multiple image error such as focal length value error and lens distortion.Its measuring principle is visual imaging, and, as long as hot spot does not depart from screen,, within the visual field of imaging system, system just can detect the position of hot spot, is not subject to the impact of spot motion speed.According to optical imaging concept, if there are two hot spots on screen simultaneously, it is the positional information that comprises two hot spots in Polaroid gradation data, can record the position coordinates of two hot spots by improving of algorithm simultaneously, and then complete position probing and the tracking to multiple target hot spots simultaneously.The detection target of measuring system is the hot spot on screen, and this hot spot is sent by the generating laser of specifying, and is incident upon and on diffuse reflection screen, forms hot spot to be detected.In practical application, can according to demand, generating laser be fixed in different targets to be measured, as turntable, transit, guide rail etc., and then complete the isoparametric measurement of corner accuracy, linearity to target to be measured.Two generating lasers can also be placed in two targets to be measured, control two targets to be measured and carry out identical action, detection system detects the track that simultaneously detects the hot spot that two targets to be measured send, by the comparison of track, and the kinematic accuracy of two targets to be measured of assessment.

Claims (6)

1. a displacement of targets device for measuring vector quantity of demarcating based on laser tracker pointwise, it is characterized in that, it comprises high speed CCD camera (1), laser tracker (2) and diffuse reflection projection screen (3), and object is upper mobile at diffuse reflection projection screen (3);

High speed CCD camera (1) is for the object calibration point on diffuse reflection projection screen (3) is carried out to imaging, and laser tracker (2) is for measuring the object calibration point coordinate on diffuse reflection projection screen (3).

2. the displacement of targets device for measuring vector quantity of demarcating based on laser tracker pointwise according to claim 1, it is characterized in that, the square of described diffuse reflection projection screen (3) for being spliced by the ground glass of two 1000mm × 500mm × 3mm, the back spraying coating black paint vehicle of this diffuse reflection projection screen (3).

3. a measuring method for the displacement of targets device for measuring vector quantity based on demarcating based on laser tracker pointwise described in claim 1, is characterized in that, it comprises the following steps:

Step 1: set upper 1 A1 of diffuse reflection projection screen (3) as object calibration point, obtaining the coordinate figure of this object calibration point A1 on projection screen coordinate system (4) by laser tracker (2) is A1 (U1, V1);

Step 2: keep object calibration point A1 to fix, utilize high speed CCD camera (1) to carry out imaging to this object calibration point A1, in the image planes of high speed CCD camera (1), obtain the corresponding conjugate points B1 of object calibration point A1, conjugate points B1 is carried out after image processing, obtain the coordinate figure B1 (m1, n1) of conjugate points B1 in image coordinates system (5);

Step 3: on diffuse reflection projection screen (3), taking object calibration point A1 as benchmark, selected 24 object calibration point Ai again, i=2,3,4,, 25,25 object calibration points are above evenly arranged with the spacing of 25cm at diffuse reflection projection screen (3), then successively according to the method for step 1 and step 2, until complete the measuring and calculating to the 25th object calibration point, obtain Ai (Ui, Vi) and Bi (mi, ni) corresponding relation look-up table, i=1,2,3,, 25;

Step 4: the screen hot spot imaging that adopts high speed CCD camera (1) successively object to be obtained in the upper projection of diffuse reflection projection screen (3), the conjugate points Bj (mj, nj) of acquisition object subpoint, j=1,2,3,

Step 5: search conjugate points Bj (mj, nj) in described corresponding relation look-up table, if Bj (mj, nj) overlaps with certain the some Bi (mi, ni) in corresponding relation look-up table, perform step six; Otherwise execution step seven;

Step 6: obtain the coordinate figure of object subpoint on projection screen coordinate system (4) according to corresponding relation look-up table, then perform step eight;

Step 7: in described corresponding relation look-up table according to Bj (mj, nj) immediate four known calibration point Bi (mi around, ni), adopt the calculation method of bilinear interpolation to obtain Aj (Uj, Vj), Aj (Uj, Vj) is the object subpoint corresponding with conjugate points Bj (mj, the nj) coordinate figure on projection screen coordinate system (4); Then perform step eight;

Step 8: initial value and final value by step 4 to the Aj (Uj, Vj) obtaining in step 7 process, calculate and obtain the displacement vector of object on diffuse reflection projection screen (3).

4. the measuring method of the displacement of targets device for measuring vector quantity of demarcating based on laser tracker pointwise according to claim 3, is characterized in that,

Described object calibration point is hot spot or target ball.

5. the measuring method of the displacement of targets device for measuring vector quantity of demarcating based on laser tracker pointwise according to claim 4, it is characterized in that, described conjugate points B1 is carried out after image processing, obtain the coordinate figure B1 (m1 of conjugate points B1 in image coordinates system (5), n1) concrete grammar is: the 8 bit image gradation datas of the conjugate points B1 first high speed CCD camera (1) being obtained carry out dual threshold pixel extraction, in the effective pixel area of conjugate points B1, utilize the method for linear interpolation to increase valid pixel number, the barycenter extraction algorithm of recycling based on intensity-weighted calculates and obtains the coordinate figure B1 (m1 of conjugate points B1 in image coordinates system (5), n1).

6. the measuring method of the displacement of targets device for measuring vector quantity of demarcating based on laser tracker pointwise according to claim 5, it is characterized in that, in the effective pixel area of conjugate points B1, utilize the method for linear interpolation to increase valid pixel number, the detailed process that the barycenter extraction algorithm of recycling based on intensity-weighted calculates the coordinate figure B1 (m1, n1) of acquisition conjugate points B1 in image coordinates system (5) is:

Below taking hypothesis a bit as example, this detailed process is described:

Pre-set the gradation of image database of conjugate points, this database comprises the multiple pixel f (p that are matrix arrangement, q), p is the line number of matrix, q is matrix column number, the 8 bit image gradation datas of the conjugate points B1 that setting high-speed ccd video camera (1) obtains are f (u, v), and these gradation of image data f (u, v) are positioned at pixel f (2,2), the f (2 of gradation of image database, 3), f (3,2) and between f (3,3)

Utilize the method acquisition f (u, 2) of linear interpolation to be:

f(u,2)=f(2,2)+(u-2)·[f(3,2)-f(2,2)]，

Obtaining f (u, 3) by f (2,3) and f (3,3) interpolation is again:

f(u,3)=f(2,3)+(u-2)·[f(3,3)-f(2,3)]，

Finally, utilize f (u, 2) and f (u, 3) to carry out interpolation, obtain f (u, v):

f(u,v)=f(u,2)+(v-2)·[f(u,3)-f(u,2)]；

The pixel f (2 of former formation hot spot, 2), f (2,3), f (3,2) and f (3,3) are by after interpolation, become f (2,2), f (2,3), f (3,2), f (3,3) and f (u, v);

Based on this, the barycenter extraction algorithm of utilization based on intensity-weighted calculates the coordinate figure B1 (m1, n1) of acquisition conjugate points B1 in image coordinates system (5) and is:

$m1=\frac{2\&CenterDot;f\left(\mathrm{2,2}\right)+2\&CenterDot;f\left(\mathrm{2,3}\right)+3\&CenterDot;f\left(\mathrm{3,2}\right)+3\&CenterDot;f\left(\mathrm{3,3}\right)+u\&CenterDot;f(u,v)}{f\left(\mathrm{2,2}\right)+f\left(\mathrm{2,3}\right)+f\left(\mathrm{3,2}\right)+f\left(\mathrm{3,3}\right)+f(u,v)},$

$n1=\frac{2\&CenterDot;f\left(\mathrm{2,2}\right)+3\&CenterDot;f\left(\mathrm{2,3}\right)+2\&CenterDot;f\left(\mathrm{3,2}\right)+3\&CenterDot;f\left(\mathrm{3,3}\right)+v\&CenterDot;f(u,v)}{f\left(\mathrm{2,2}\right)+f\left(\mathrm{2,3}\right)+f\left(\mathrm{3,2}\right)+f\left(\mathrm{3,3}\right)+f(u,v)}.$

Images