CN104976968A - Three-dimensional geometrical measurement method and three-dimensional geometrical measurement system based on LED tag tracking - Google Patents

Abstract

The invention provides a three-dimensional geometrical measurement method and a three-dimensional geometrical measurement system based on LED tag tracking. A three-dimensional scanner is arranged such that the three-dimensional scanner can move along a track and is close to an object to be measured, and blocked local measurement is flexibly and conveniently carried out on a large-scale object by a structured light method to obtain a high-resolution local image. Meanwhile, as the light of an LED tag is bright enough, even when the distance between a stereoscopic tracker and the LED tag increases, the stereoscopic tracker can quickly and accurately track the LED tag to ensure that a server can precisely locate the center position of the LED tag and the position of the LED tag in the coordinate system of the three-dimensional scanner, precisely obtain the LED tag relationship between two local measurements, accurately evaluate the attitude change of the three-dimensional scanner and realize precise stitching of local measurement results. Moreover, three-dimensional measurement of the whole contour of a large-scale object can be completed conveniently, stably and accurately without the need for tedious switching between multiple projectors.

Description

A kind of three-dimensional geometry measuring method based on the tracking of LED label and system

Technical field

The present invention relates to three-dimensional measurement technical field, particularly relate to a kind of three-dimensional geometry measuring method based on the tracking of LED label and system.

Background technology

At present, three-dimensional measurement technology mainly comprises contact type measurement and non-contact measurement two class.In commercial production, some large scale workpiece can only adopt non-contact 3-D measuring technique, as ship surface steel plate.Non-contact 3-D measuring technique comprises two classes: vision photograph three-dimensional measurement, 3-d laser measurement.Photograph three-dimension measuring system has the advantage that measuring speed is fast, density measurement is high.Laser measurement system has the high advantage of measuring accuracy, but when measuring large scale target, laser scanning speed is comparatively slow, cannot reach the requirement of real-time in commercial production.

Vision measurement mainly can be divided into: passive measurement and Active measuring two kinds of methods.Passive measurement method realizes three-dimensional modeling by the surperficial light sent or reflect of detecting object.But the method needs body surface to have abundant texture structure, therefore, be difficult to apply it in commercial production.Active measuring is by machinery or radiation mode contact testee, and existing a lot of Active measuring method, by projecting specific light to testee, measures object three-dimensional form.Compared with passive measurement method, Active measuring is more suitable for commercial production, because they can be more stable, and obtains denser data.

Active measuring method probably can be divided into two classes: time-of-flight (TOF) laser measurement and structural light measurement.(1) TOF laser measurement method obtains the 3D shape of object by the flight time calculating light.(2) structural light measurement is projected on testee by projector by the light of coding, and camera is caught and rebuild these scenes simultaneously.Use TOF camera can obtain dense 3-D view in real time.But TOF camera resolution is too low, and random noise in depth map is too large, and the three-dimensional image quality therefore obtained is not high, and therefore, comparatively speaking, structural light measurement more meets the requirement of commercial production to precision.

Large scale object once cannot complete measurement, needs first local measurement, then splices.Such as a kind of large-scale steel plate three-dimensional measurement splicing system and method disclosed in application number CN201310358478.7, first utilize two background plane instrument to the texture of tested steel plate projection complexity, and the texture picture as a setting of steel plate is taken with the camera of two in spatial digitizer, then two background plane instrument are closed, open the projector of spatial digitizer, with two camera shooting steel plate images, the three-dimensional data of the steel plate taken by server obtains, and adopt SIFT algorithm and RANSAC method to process the background picture of each several part steel plate and three-dimensional data, to splice the three-dimensional data of adjacent part steel plate block, obtain the three-dimensional data of whole steel plate thus.This large-scale steel plate three-dimensional measurement splicing system and method, background plane instrument needs to be erected at height in the air on the one hand, to ensure to steel plate whole surface projection special texture, cause projector distance surface of steel plate far away thus, divergence of beam and dying down, the light that arrival surface of steel plate reflects into spatial digitizer becomes more weak, in this case the background image that spatial digitizer gathers can be clear not, make troubles to later image process and data analysis, and then affect the three-dimensional data of steel plate and the precision of image mosaic; All need when each local measurement on the other hand to switch between two projector of peripheral hardware and the built-in projector of spatial digitizer, cause whole measurement complex operation, system uses quite inconvenient.

Therefore, need a kind of not by natural lighting variable effect and can easily and fast, exactly large scale object is carried out to the method and system of vision measurement.

Summary of the invention

The object of the present invention is to provide a kind of large scale three-dimensional measurement of objects method and system of following the tracks of based on LED label, vision measurement can be carried out to large scale object easily and fast, exactly, eliminate natural light Strength Changes to the adverse effect of measuring accuracy simultaneously.

For solving the problem, the present invention proposes a kind of three-dimensional geometry measuring method of following the tracks of based on LED label, comprising:

The first step, being arranged on by a set of spatial digitizer posting LED label can along on the support of the rail moving of object under test side, and by described support-moving one end to described track, described spatial digitizer mainly contains and is made up of two industrial cameras and a projector;

Second step, is fixedly installed three-dimensional tracking instrument, and the fixed position of three-dimensional tracking instrument makes can trace into described LED label all the time at measuring process neutral body tracker;

3rd step, start a local measurement, the projector of described spatial digitizer is to object under test surface projective structure light, two industrial cameras of described spatial digitizer take the object under test topography in the respective visual field, and described partial image data is transferred to server, meanwhile, three-dimensional tracking instrument is taken LED label image and is transferred to server;

4th step, described server generates the 3-D view of a local of described object under test according to the partial image data of described two industrial cameras, in described LED label image, accurately locate LED tag hub point position simultaneously, complete the measurement of a local of described object under test;

5th step, promotes support-moving along track and changes to reposition to promote spatial digitizer, repeats third and fourth step, completes the measurement of the next local of object under test;

6th step, described server adopts KLT algorithm, the LED label relation between the LED label image setting up two local measurements;

7th step, described server obtains the LED label position in the spatial digitizer coordinate system of each local measurement according to the LED label image of each local measurement;

8th step, described server according to the described LED label relation of two local measurements with the described LED label position of two local measurements, obtain the change of the described LED label position of described two local measurements, to assess the spatial digitizer attitudes vibration of described two local measurements;

9th step, described server, according to the result of the 8th step, splices the 3-D view of described two local measurements;

Tenth step, repeats step 5 to nine, until last local measurement of described object under test terminates, to splice the 3-D view on the whole surface of described object under test.

Further, described spatial digitizer mainly contains and is made up of two industrial cameras and a projector.

Further, the luminous power of described LED label is more than or equal to 3W.

Further, described three-dimensional tracking instrument forms primarily of two industrial cameras.

Further, the camera resolution of described spatial digitizer and described three-dimensional tracking instrument all more than 1440 × 1080, frame per second all at more than 10fps, and respectively by kilomega network connection server; The projector of described spatial digitizer is by USB line connection server.

Further, in described 4th step, in described LED label image, accurately the step of LED tag hub point position, location comprises:

2a: adopt normalization template matching method, find the LED label image block in described LED label image;

2b: amplify described LED label image block and obtain enlarged image Pm, and in 3-D view corresponding to Pm, LED label Lm is rendered as octagonal mountain shape;

2c: adopt Gaussian function to the smoothing process of the 3-D view that Pm is corresponding, obtain a smoothed image, and determine the summit Pt of Lm image according to smoothed image summit;

2d: face based on 3-D view PV: one bottom surface Pb according to described level and smooth Computer image genration with five virtual planes, four cross facets all intersect at same intersection L perpendicular to Pb, and the angle between adjacent two crossing plane is 45 degree;

2e: superimposed images Lm and PV, and the bottom of Lm and PV overlaps and allows L by the summit Pt of Lm image;

2f: allow four cross facets of PV take L as axle, entirety rotates predetermined angle in the direction of the clock successively until rotate a circle, and records the intersection cross-sectional area of each postrotational PV and Lm;

2g: the current location of L and maximum intersection cross-sectional area part are recorded in queue Qs;

2h: 24 neighbor pixels representing Pt with Npt, mobile PV, allow L by each pixel in Npt, repeats step 2f ~ 2h;

2i: in Qs, selects the element with maximum intersection cross-sectional area, selects the position of the L that described element is corresponding as true center Ca;

2j: reduce the multiplying power of Pm to described LED label image block, and obtain the LED tag hub point position in described LED label image according to Ca.

Further, in described 7th step, described server comprises the step that the LED label position in the spatial digitizer coordinate system of each local measurement is demarcated:

3a: a gridiron pattern scaling board is placed in the visual field of three-dimensional tracking instrument and spatial digitizer, three-dimensional tracking instrument can also trace into the LED label on spatial digitizer simultaneously;

3b: obtain the transformational relation between three-dimensional tracking instrument coordinate system and spatial digitizer coordinate system by described gridiron pattern scaling board;

3c: the position of LED label in three-dimensional tracking instrument coordinate system being obtained each local measurement by the LED label image of three-dimensional tracking instrument shooting during each local measurement, and the position of LED label in three-dimensional tracking instrument coordinate system is converted to the position of LED label in spatial digitizer coordinate system by described transformational relation.

Further, in described 8th step, according to the change of the described LED label position of described two local measurements, assess the spatial digitizer attitudes vibration of described two local measurements, comprising:

4a: by the position T of spatial digitizer during certain local measurement ^tbe expressed as vectorial Y, wherein,

$T^t=\left(\begin{array}{cccc}{T}_{11}^t& T_{12}^t& T_{13}^t& T_{14}^t\\ T_{21}^t& T_{22}^t& T_{23}^t& T_{24}^t\\ T_{31}^t& T_{32}^t& T_{33}^t& T_{34}^t\\ 0& 0& 0& 1\end{array}\right),Y=(T_{11}^T,T_{12}^T,T_{13}^T,T_{14}^T,T_{21}^T,T_{23}^T,T_{24}^T,T_{31}^T,T_{32}^T,T_{33}^T,T_{34}^T);$

4b: described in definition, during certain local measurement, the left and right projection matrix of three-dimensional tracking instrument is respectively:

$P_l=\left(\begin{array}{cccc}{P}_l^{11}& P_l^{12}& P_l^{13}& P_l^{14}\\ P_l^{21}& P_l^{22}& P_l^{23}& P_l^{24}\\ P_l^{31}& P_l^{32}& P_l^{33}& P_l^{34}\end{array}\right);P_r=\left(\begin{array}{cccc}{P}_r^{11}& P_r^{12}& P_r^{13}& P_r^{14}\\ P_r^{21}& P_r^{22}& P_r^{23}& P_r^{24}\\ P_r^{31}& P_r^{32}& P_r^{33}& P_r^{34}\end{array}\right);$

4c: according to P _l, P _rand T ^tby the coordinate X of LED label in spatial digitizer coordinate system during certain local measurement described _s ⁱproject in three-dimensional tracking instrument, set up projection relation:

$\left\{\begin{array}{c}{P}_l{T}^t{X}_S^i=x_l^i\\ P_r{T}^t{X}_S^i=x_r^i\end{array}\right.,i\∈1,\·\·\·,N;$

Wherein, with the pixel coordinate of LED label respectively in three-dimensional tracking instrument middle left and right two width image; the coordinate of LED label in spatial digitizer coordinate system;

4d: according to described projection relation and T ^tand the relation between Y obtain vectorial Y with p _l, P _rbetween least square method matrix relationship AY=b, wherein,

$A=\left(\begin{array}{c}(P_l^{11}-x_l^{i_{\mathrm{\λ}}}{P}_l^{31})x_S^{i_{\mathrm{\λ}}},(P_l^{11}-x_l^{i_{\mathrm{\λ}}}{P}_l^{31})x_S^{i_{\mathrm{\λ}}},(P_l^{11}-x_l^{i_{\mathrm{\λ}}}{P}_l^{31})x_S^{i_c},(P_l^{11}-x_l^{i_{\mathrm{\λ}}}{P}_l^{31}),(P_l^{12}-x_l^{i_{\mathrm{\λ}}}{P}_l^{32})x_S^{i_{\mathrm{\λ}}},(P_l^{12}-x_l^{i_{\mathrm{\λ}}}{P}_l^{32})x_S^{i_y},(P_l^{12}-x_l^{i_{\mathrm{\λ}}}{P}_l^{32})x_S^{i_c},(P_l^{12}-x_l^{i_{\mathrm{\λ}}}{P}_l^{32}),(P_l^{13}-x_l^{i_{\mathrm{\λ}}}{P}_l^{33})x_S^{i_{\mathrm{\λ}}},(P_l^{13}-x_l^{i_{\mathrm{\λ}}}{P}_l^{33})x_S^{i_y},(P_l^{13}-x_l^{i_{\mathrm{\λ}}}{P}_l^{33})x_S^{i_c}(P_l^{13}-x_l^{i_{\mathrm{\λ}}}{P}_l^{33})\\ (P_l^{21}-x_l^{i_{\mathrm{\λ}}}{P}_l^{31})x_S^{i_{\mathrm{\λ}}},(P_l^{21}-x_l^{i_{\mathrm{\λ}}}{P}_l^{31})x_S^{i_{\mathrm{\λ}}},(P_l^{21}-x_l^{i_{\mathrm{\λ}}}{P}_l^{31})x_S^{i_c},(P_l^{21}-x_l^{i_{\mathrm{\λ}}}{P}_l^{31}),(P_l^{22}-x_l^{i_{\mathrm{\λ}}}{P}_l^{32})x_S^{i_{\mathrm{\λ}}},(P_l^{22}-x_l^{i_{\mathrm{\λ}}}{P}_l^{32})x_S^{i_y},(P_l^{22}-x_l^{i_{\mathrm{\λ}}}{P}_l^{32})x_S^{i_c},(P_l^{22}-x_l^{i_{\mathrm{\λ}}}{P}_l^{32}),(P_l^{23}-x_l^{i_{\mathrm{\λ}}}{P}_l^{33})x_S^{i_{\mathrm{\λ}}},(P_l^{23}-x_l^{i_{\mathrm{\λ}}}{P}_l^{33})x_S^{i_y},(P_l^{23}-x_l^{i_{\mathrm{\λ}}}{P}_l^{33})x_S^{i_c},(P_l^{23}-x_l^{i_{\mathrm{\λ}}}{P}_l^{33})\\ (P_r^{11}-x_r^{i_{\mathrm{\λ}}}{P}_r^{31})x_S^{i_{\mathrm{\λ}}},(P_r^{11}-x_r^{i_{\mathrm{\λ}}}{P}_r^{31})x_S^{i_y},(P_r^{11}-x_r^{i_{\mathrm{\λ}}}{P}_r^{31})x_S^{i_c},(P_r^{11}-x_r^{i_{\mathrm{\λ}}}{P}_r^{31}),(P_r^{12}-x_r^{i_{\mathrm{\λ}}}{P}_r^{32})x_S^{i_{\mathrm{\λ}}},(P_r^{12}-x_r^{i_{\mathrm{\λ}}}{P}_r^{32})x_S^{i_y},(P_r^{12}-x_r^{i_{\mathrm{\λ}}}{P}_r^{32})x_S^{i_c},(P_r^{12},x_r^{i_{\mathrm{\λ}}}{P}_r^{32}),(P_r^{13}-x_r^{i_{\mathrm{\λ}}}{P}_r^{33})x_S^{i_{\mathrm{\λ}}},(P_r^{13}-x_r^{i_{\mathrm{\λ}}}{P}_r^{33})x_S^{i_y},(P_r^{13}-x_r^{i_{\mathrm{\λ}}}{P}_r^{33})x_S^{i_c},(P_r^{13}-x_r^{i_{\mathrm{\λ}}}{P}_r^{33})\\ (P_r^{21}-x_r^{i_{\mathrm{\λ}}}{P}_r^{31})x_S^{i_{\mathrm{\λ}}},(P_r^{21}-x_r^{i_{\mathrm{\λ}}}{P}_r^{31})x_S^{i_y},(P_r^{21}-x_r^{i_{\mathrm{\λ}}}{P}_r^{31})x_S^{i_c},(P_r^{21}-x_r^{i_{\mathrm{\λ}}}{P}_r^{31}),(P_r^{22}-x_r^{i_{\mathrm{\λ}}}{P}_r^{32})x_S^{i_{\mathrm{\λ}}},(P_r^{22}-x_r^{i_{\mathrm{\λ}}}{P}_r^{32})x_S^{i_y},(P_r^{22}-x_r^{i_{\mathrm{\λ}}}{P}_r^{32})x_S^{i_c},(P_r^{22}-x_r^{i_{\mathrm{\λ}}}{P}_r^{33})x_S^{i_{\mathrm{\λ}}},(P_r^{23}-x_r^{i_{\mathrm{\λ}}}{P}_r^{33})x_S^{i_y},(P_r^{23}-x_r^{i_{\mathrm{\λ}}}{P}_r^{33})x_S^{i_c},(P_r^{23}-x_r^{i_{\mathrm{\λ}}}{P}_r^{33})\end{array}\right)$ ； $b=\left(\begin{array}{c}(x_l^{i_x}{P}_l^{34}-P_l^{14})\\ (x_l^{i_y}{P}_l^{34}-P_l^{24)}\\ (x_r^{i_x}{P}_r^{34}-P_r^{14})\\ (x_r^{i_x}{P}_r^{34}-P_r^{24})\end{array}\right);$

4e: the attitude calculating the spatial digitizer of certain local measurement described according to described least square method matrix relationship is: Y ^*=argmin _y(| AY-b| ²);

4f: the spatial digitizer attitudes vibration assessing two local measurements according to the attitude of the spatial digitizer calculated during double local measurement.

The present invention also provides a kind of three-dimensional geometry measuring system of following the tracks of based on LED label applying above-mentioned three-dimensional geometry measuring method, comprises a station server, a set of three-dimensional tracking instrument and a set of spatial digitizer posting LED label; Wherein, described spatial digitizer is arranged on can along on the support of the rail moving of object under test side, mainly contains and can be made up of to the projector of described object under test surface projective structure light two industrial cameras can taking the topography of object under test and one; Three-dimensional tracking instrument is fixedly installed and its fixed position makes can photograph described LED label image all the time at measuring process neutral body tracker; Two industrial cameras of three-dimensional tracking instrument and spatial digitizer are respectively by kilomega network connection server, and projector is by USB line connection server; Described server comprises:

According to the partial 3 d image composing unit of topography's synthesis partial 3 d image of two industrial camera shootings;

The KLT algorithm unit of the LED label relation between the LED label image that the three-dimensional tracking instrument setting up two local measurements is taken;

The LED obtaining the LED label position in the spatial digitizer coordinate system of each local measurement according to the LED label image of each local measurement demarcates unit;

The attitude assessment unit of the change obtaining the described LED label position of described two local measurements and the spatial digitizer attitudes vibration assessing described two local measurements;

Splice the image mosaic unit of the 3-D view of continuous two local measurements.

Compared with prior art, three-dimensional geometry measuring method based on the tracking of LED label provided by the invention and system, when carrying out vision measurement to large scale object, have employed in fact method of structured light and local measurement is carried out to large scale object, then adopting the method splicing local measurement result based on following the tracks of, achieving following technique effect thus:

1, spatial digitizer is configured to track-movable device, can flexibly, easily piecemeal measure large sized object, simultaneously its projector and two industrial cameras all nearer apart from local surfaces to be measured when each local measurement, projector can project the textured pattern of high-resolution to local surfaces to be measured and the light reflexed in two industrial cameras is stronger, so two industrial cameras can photograph the topography of high definition, avoid the unnecessary trouble in later image process and data analysis, improve the three-dimensional data of steel plate and the precision of image mosaic;

2, the present invention arranges LED label as splicing mark on spatial digitizer, because the luminosity of LED label is enough strong, so after the movement of spatial digitizer, especially when the distance of three-dimensional tracking instrument and LED label increases, three-dimensional tracking instrument can be light, promptly follow the tracks of the high-brightness LED label on spatial digitizer, make three-dimensional tracking instrument can ignore LED label surrounding environment and shoot high definition simultaneously, the LED label image of dispersion shape, thus make server accurately can orient LED tag hub position and the LED label relation obtained between twice local measurement and the positional information of LED label in spatial digitizer coordinate system, and then can the attitudes vibration of accurate evaluation spatial digitizer, realize the accurate splicing to local measurement result, therefore without the need to carrying out the loaded down with trivial details switching between multiple projector, just can be flexible, convenient, stable, accurately complete the three-dimensional measurement to the whole profile of large scale object,

3, the center position of anistree mountain shape LED label that the present invention is further advanced by the LED label image that structure five virtual planes take three-dimensional tracking instrument is accurately located, and further increases splicing precision.

Accompanying drawing explanation

Fig. 1 is the configuration diagram of the three-dimensional geometry measuring system based on the tracking of LED label of the specific embodiment of the invention;

Fig. 2 is the system architecture schematic diagram of the server of the specific embodiment of the invention;

Fig. 3 is the process flow diagram of the three-dimensional geometry measuring method based on the tracking of LED label of the specific embodiment of the invention;

Fig. 4 is that schematic diagram is detected in the loose shape LED label true center position in method shown in Fig. 3;

Fig. 5 is the position view of calculating LED label in spatial digitizer coordinate system in method shown in Fig. 3.

Embodiment

For making object of the present invention, feature becomes apparent, and be further described, but the present invention can realize by different forms, should just not be confined to described embodiment below in conjunction with accompanying drawing to the specific embodiment of the present invention.

Please refer to Fig. 1, the invention provides a kind of three-dimensional geometry measuring system of following the tracks of based on LED label, comprise a high performance server 40, a set of three-dimensional tracking instrument 30 and a set of spatial digitizer 20 posting the LED label 23 of high brightness.Wherein, described spatial digitizer 20 is arranged on can along on the support 12 of track 11 movement of object under test 10 side, mainly contains and can be made up of to the projector 21 of the surperficial projective structure light of described object under test 10 two industrial cameras 22 can taking the topography of object under test 10 and one.Preferably, object under test is placed on a tester table, and track 11 is arranged at the both sides of tester table.Three-dimensional tracking instrument 30 is fixedly installed and its fixed position makes can photograph all the time at measuring process neutral body tracker 30 image of described LED label 23, in the present embodiment, three-dimensional tracking instrument forms primarily of two industrial cameras 31, be arranged on the outside of track end, before starting measurement, spatial digitizer 20 and three-dimensional tracking instrument 30 are lived apart the two ends of track, and during measurement, spatial digitizer 20 can draw near close to three-dimensional tracking instrument 30 along track.Certainly, also near three-dimensional tracking instrument 30, during measurement, from the close-by examples to those far off each local measurement can be carried out successively along track when spatial digitizer 20 is initial.And the resolution of the industrial camera of three-dimensional tracking instrument 30 and spatial digitizer 20 is all more than 1440 × 1080, frame per second all at more than 10fps, and respectively by the network interface of kilomega network connection server 40; The projector 21 of spatial digitizer 20 passes through the USB interface of USB line connection server.The luminous power of preferred LED label is at more than 3W.

Please refer to Fig. 2, the server of the present embodiment comprises partial 3 d image composing unit 401, KLT algorithm unit 402, LED demarcation unit 403, attitude assessment unit 404, image mosaic unit 405 and network interface and USB interface.

Wherein, the visual field of two industrial cameras 22 of spatial digitizer 20 cannot the global image of disposable shooting large scale object under test, local gradation can only be carried out measure, therefore spatial digitizer 20 is without the need to being arranged on very high position, projector 21 can in the low empty position projective structure light (special texture of the local location of object under test 10, be such as grating pattern), incident light is stronger, the relative grow of reflected light, two industrial cameras 22 of spatial digitizer 20 are also at the low empty position of the local location of object under test 10 simultaneously, so the picture rich in detail of the local one by one of large scale object under test can be photographed.The network interface 408 that partial image data transfers to server 40 by kilomega network receives after taking by two industrial cameras 22.Simultaneously, because the lamp of LED label is according to very bright, three-dimensional tracking instrument 30 can ignore LED label surrounding environment easily, focuses on LED label itself, and photograph the high-definition image of LED label on spatial digitizer 20, and received by the network interface 407 that kilomega network transfers to server 40.

Partial 3 d image composing unit 401 is for synthesizing a partial 3 d image by the left and right local two dimensional image of two of spatial digitizer 20 objects under test that industrial camera 22 is taken.In the present embodiment, two industrial cameras 31 of three-dimensional tracking instrument 30 can photograph the left images of LED label 23, therefore partial 3 d image composing unit 401 also for the synthesis of LED label 23 3-D view and determine the center of the LED label in LED label image.Because the lamp of LED label is according to very bright, the LED label image of two industrial camera shootings of three-dimensional tracking instrument 30 is all dispersion shapes, instead of circular, and the 3-D view of the LED label synthesized thus is rendered as octagonal mountain shape.

KLT algorithm unit 402 for receiving the LED label image of three-dimensional tracking instrument shooting, and adopts KLT algorithm to set up LED label relation between the LED label image of two local measurements.KLT algorithm be a kind of with window W to be tracked video image interframe gray scale difference quadratic sum (SSD, Sum of Squared intensity Differences) as tolerance track algorithm.Clear picture in order to make spatial digitizer 20 and three-dimensional tracking instrument 30 photograph in the present embodiment, can arrange the stand-by period takes image, namely change to new position when spatial digitizer and just gather image after a period of stabilisation, and the LED label image that three-dimensional tracking instrument 30 corresponding to spatial digitizer 20 position is taken is a frame, therefore KLT algorithm unit 402 is in fact the relation setting up LED label between continuous print LED label image frame.

Because two of spatial digitizer 20 industrial cameras 22 and LED label 23 are in same plane, so the image of LED label directly cannot be obtained, namely the position of LED label directly can not be obtained in the coordinate system of spatial digitizer 20, need the position of locating LED label in spatial digitizer coordinate system by server 40, demarcate also referred to as LED.LED demarcates the LED label image of unit 403 namely for taking when each local measurement according to three-dimensional tracking instrument 30, obtains the LED label position in the spatial digitizer coordinate system of each local measurement.

Attitude assessment unit 404 is for obtaining the change of the LED label position of described two local measurements and assessing the spatial digitizer attitudes vibration of two local measurements, namely the change of the LED label position of two local measurements is first assessed, the attitudes vibration of the spatial digitizer 20 when assessing two local measurements again, assesses also referred to as spatial digitizer attitude.

The LED label relation that image mosaic unit 405 obtains for the attitudes vibration of spatial digitizer 20 that evaluates according to attitude assessment unit 404 and KLT algorithm unit 402, splice the 3-D view of two local measurements, thus splice the 3-D view of the whole contour surface of object under test.In the present embodiment, because spatial digitizer 20 is along rail moving, therefore former and later two local measurements are continuous print, image mosaic unit 405 adopts the mode of splicing in turn, namely the back segment edge of the front edge of a rear partial 3 d image and previous partial 3 d image can be stitched together, all local measurements (such as the 1 to the i-th-1 local measurement above will be gone through thus, i is greater than 2) and splice the 3-D view (1+2+ ... + i-1) splice with the 3-D view of a rear local measurement (i-th local), form integrating three-dimensional image (1+2+ ... + i-1+i), and then complete the splicing of whole profile.

Please refer to Fig. 3, the present invention proposes a kind of three-dimensional geometry measuring method of following the tracks of based on LED label, comprising:

The first step, being arranged on by a set of spatial digitizer posting LED label can along on the support of the rail moving of object under test side, and by described support-moving one end to described track, described spatial digitizer mainly contains and is made up of two industrial cameras and a projector;

Second step, is fixedly installed three-dimensional tracking instrument, and the fixed position of three-dimensional tracking instrument makes can trace into described LED label all the time at measuring process neutral body tracker;

3rd step, start a local measurement, the projector of described spatial digitizer is to object under test surface projective structure light, two industrial cameras of described spatial digitizer take the object under test topography in the respective visual field, and described partial image data is transferred to server, meanwhile, three-dimensional tracking instrument is taken LED label image and is transferred to server;

4th step, described server generates the 3-D view of a local of described object under test according to the partial image data of described two industrial cameras, in described LED label image, accurately locate LED tag hub point position simultaneously, complete the measurement of a local of described object under test;

5th step, promotes support-moving along track and changes to reposition to promote spatial digitizer, repeats third and fourth step, completes the measurement of the next local of object under test;

6th step, described server adopts KLT algorithm, the LED label relation between the LED label image setting up two local measurements;

7th step, described server obtains the LED label position in the spatial digitizer coordinate system of each local measurement according to the LED label image of each local measurement;

8th step, described server according to the described LED label relation of two local measurements with the described LED label position of two local measurements, obtain the change of the described LED label position of described two local measurements, to assess the spatial digitizer attitudes vibration of described two local measurements;

9th step, described server, according to the result of the 8th step, splices the 3-D view of described two local measurements;

Tenth step, repeats step 5 to nine, until last local measurement of described object under test terminates, to splice the 3-D view on the whole surface of described object under test.

Below with large scale plate with curved surface 10 for measuring object, composition graphs 1 to 5 describe in detail three-dimensional geometry measuring method of the present invention.

The first step, spatial digitizer 20 is installed: be arranged on by spatial digitizer 20 on support 12, with along track 11 movement flexibly, with flexibly, easily piecemeal measure large scale plate with curved surface 10, all be connected on the network interface 408 of server 40 via kilomega network by two of spatial digitizer 20 industrial cameras 22, the projector 21 in spatial digitizer 20 is by the USB interface 406 of USB line connection server 40 simultaneously.

Second step, three-dimensional tracking instrument 3 is installed: three-dimensional tracking instrument 30 is placed on fixed position, its two industrial cameras 22 are all connected on the network interface 407 of server 40 via kilomega network, ensure: in measuring process, two industrial cameras 31 of three-dimensional tracking instrument 30 can trace into the LED label 23 on spatial digitizer 20 all the time simultaneously.

3rd step, start a local measurement: energising, start a local measurement, now LED label 23, two industrial cameras 22 of spatial digitizer 20 and projector 21 all enter duty, and two industrial cameras 31 of three-dimensional tracking instrument 30 all enter duty.During each local measurement, the projector 21 of spatial digitizer 20 all projects bar shaped grating texture (i.e. structured light) to steel plate 1, two industrial cameras 22 of spatial digitizer 20 obtain the topography of the steel plate 10 in the visual field simultaneously, and be transferred to server 40 by kilomega network, two industrial cameras 31 of three-dimensional tracking instrument 30 take the image of the LED label 23 on spatial digitizer 20, and are transferred to server 40 by kilomega network.

4th step, the left and right topography that two of spatial digitizer 20 industrial camera 22 is taken can be carried out three-dimensional synthesis by the partial 3 d image composing unit 401 of server 40, obtain the partial 3 d image of steel plate 10, accurately LED tag hub point position is located in the image of the LED label 23 simultaneously on the spatial digitizer 20 of two industrial cameras 31 shootings of three-dimensional tracking instrument 30, specific as follows:

2a: adopt normalization template matching method, in the LED label image finding two industrial cameras 31 of three-dimensional tracking instrument 30 to take, the Position Approximate of LED label, obtains the image block centered by LED label.Next for an image block, as shown in Fig. 4 (a), the true center how obtaining LED label is described.

2b: first this image block (the LED image of the dispersion shape namely shown in Fig. 4 (a)) is expanded several times, such as expand 3 times, the image block after amplification is denoted as Pm, as shown in Fig. 4 (b).Image block after amplification can synthesize the 3-D view of LED label, as shown in Fig. 4 (c), x, y coordinate of the 3-D view of LED label is the pixel coordinate of Fig. 4 (b), z coordinate is the gray-scale value of respective pixel, the 3D shape that the LED label of two-dimensional scattering shape is corresponding is rendered as octagonal mountain shape, octagonal eight angles constitute eight ridges on mountain, and this shape is called for short anistree mountain, is denoted as Lm.

2c: adopt Gaussian function to the smoothing process of Pm, obtain a level and smooth image, Fig. 4 (d) is the three dimensional form of sharpening result.The window of Gaussian smoothing is set to 2/3 of image block width.As can be seen from Fig. 4 (d), the maximal value Pt of Lm can be found easily.But because the noise existed, so Pt may not be the true center position of LED label.

2d: in order to obtain tag hub more accurately, generates five virtual planes, is denoted as PV, as shown in Fig. 4 (e).Wherein, face based on a plane, is denoted as Pb.Other four cross facets all perpendicular to Pb, and intersect at same intersection L, and the angle between adjacent two crossing plane is 45 degree.

2e: Fig. 4 (c) and Fig. 4 (e) is overlapping, and allow L pass through a Pt, and allow the bottom of two 3D shapes overlap, result is as shown in Fig. 4 (f).

2f: allow four cross facets integrally in the direction of the clock take L as axle, often rotates 1 degree with regard to the intersect cross-sectional area of record PV with 3D shape Lm, until four cross facets rotate a circle around L axle.

2g: the position (2f process for the first time, the position of L is Pt, is the respective pixel in Npt afterwards) that L is passed through now and maximum intersection cross-sectional area part are recorded in queue Qs.

2h: 24 neighbor pixels representing Pt with Npt, mobile PV, allow L by each pixel in Npt, repeats step 2e-2h, namely make L by 24 neighbor pixel points successively, carry out 2e-2h.

2i: in Qs, selects to have the element of maximum intersection cross-sectional area, selects relevant position that L passes through as true center Ca, as shown in Fig. 4 (g).

2j: reduce the Pm with true center Ca, makes Pm be contracted to the tile size of Fig. 4 (a), according to Ca, can obtain the sub-pixel center of LED label, as shown in Fig. 4 (h).

5th step, promotes spatial digitizer 20 and moves on the rail 11, repeat third and fourth step, complete the next local measurement of steel plate 10, namely measure the next local data of steel plate 1.

6th step, server 40 adopts KLT algorithm, when setting up twice local measurement the LED label 23 that two industrial cameras 31 of three-dimensional tracking instrument 30 are taken two images in the relation of LED label, particularly: by KLT feature point detection and track algorithm, defined feature match window between the two continuous frames LED label image of front and back, a large amount of character pair points is set up between two frame LED label images, obtain characteristic of correspondence point set, by these characteristic of correspondence point sets, to publish picture the affine Transform Model parameter of affined transformation (between two frames) between picture with least square fitting, a rear two field picture is mapped in previous frame image and goes, front and back two two field pictures carry out the result of time difference, thus set up the relation of LED label in two images.Further, new unique point can be supplemented, as tracking characteristics point during next local measurement, and repeat to be cycled to repeat said process.

7th step, server 40 carries out the demarcation of LED label: because in spatial digitizer 20 coordinate system, directly can not obtain the position of LED label 23 on spatial digitizer 20.Therefore, server 40 will locate the position of LED label 23 in spatial digitizer 20 coordinate system, demarcates, please refer to Fig. 5 also referred to as LED, specific as follows:

3a: a gridiron pattern scaling board 100, the place that two industrial cameras 31 of two industrial cameras 22 and three-dimensional tracking instrument 30 that are placed on spatial digitizer 20 can be seen, namely gridiron pattern scaling board 100 appears in the visual field of industrial camera 22,31 simultaneously.Meanwhile, two industrial cameras 31 of three-dimensional tracking instrument 30, also must can see the LED label 23 on spatial digitizer 20.

3b: scaling board 100 can regard world coordinate system as, is denoted as Cw; The coordinate system of spatial digitizer 20 is denoted as Cs, and the coordinate system of three-dimensional tracking instrument 30 is denoted as Ct.

The conversion of 3c:Cs to Cw is denoted as Tsw; The conversion of Ct to Cw is denoted as Ttw; The conversion of Ct to Cs is denoted as Tts.Tsw, Ttw and Tts are the matrixes that 4 row 4 arrange.In order to demarcate LED label, first Tts must be known, according to the conversion between Cw, Cs and Ct:

T _TW＝T _TST _SW(1)

Therefore, can obtain:

$T_{\mathrm{TS}}=T_{\mathrm{TW}}{T}_{\mathrm{SW}}^{-1}---\left(2\right)$

Can Tsw be obtained by scaling board, Ttw, so also just can obtain Tts.

3d: the position calculating the LED label in spatial digitizer coordinate system.Suppose that LED label 23 coordinate in Ct is can pass through formula (3) will be transformed into spatial digitizer 20 coordinate system, obtain the i.e. position of LED label 23 in spatial digitizer 20.

$X_S^i=T_{\mathrm{TS}}{X}_T^i,i\∈\{1,\·\·\·,N\}---\left(3\right)$

8th step, server 40 is according to the change in location of LED label 23, and when assessing twice local measurement, the attitudes vibration of spatial digitizer 20, assesses also referred to as spatial digitizer attitude, specific as follows:

4a. supposes t (i.e. certain local measurement moment), and the position of spatial digitizer is:

$T^t=\left(\begin{array}{cccc}{T}_{11}^t& T_{12}^t& T_{13}^t& T_{14}^t\\ T_{21}^t& T_{22}^t& T_{23}^t& T_{24}^t\\ T_{31}^t& T_{32}^t& T_{33}^t& T_{34}^t\\ 0& 0& 0& 1\end{array}\right)---\left(4\right)$

And formula (4) is rewritten into a new vectorial Y:

$Y=(T_{11}^T,T_{12}^T,T_{13}^T,T_{14}^T,T_{21}^T,T_{23}^T,T_{24}^T,T_{31}^T,T_{32}^T,T_{33}^T,T_{34}^T)---\left(5\right)$

4b. is according to the left and right projection matrix of formula (6), formula (7) definition three-dimensional tracking instrument:

$P_l=\left(\begin{array}{cccc}{P}_l^{11}& P_l^{12}& P_l^{13}& P_l^{14}\\ P_l^{21}& P_l^{22}& P_l^{23}& P_l^{24}\\ P_l^{31}& P_l^{32}& P_l^{33}& P_l^{34}\end{array}\right)---\left(6\right)$

$P_r=\left(\begin{array}{cccc}{P}_r^{11}& P_r^{12}& P_r^{13}& P_r^{14}\\ P_r^{21}& P_r^{22}& P_r^{23}& P_r^{24}\\ P_r^{31}& P_r^{32}& P_r^{33}& P_r^{34}\end{array}\right)---\left(7\right)$

4c. is according to P _l, P _rand T ^tby the coordinate projection of LED label in spatial digitizer coordinate system in three-dimensional tracking device, as shown in formula (8):

$\left\{\begin{array}{c}{P}_l{T}^t{X}_S^i=x_l^i\\ P_r{T}^t{X}_S^i=x_r^i\end{array}\right.,i\∈1,\·\·\·,N---\left(8\right)$

Wherein, with the pixel coordinate of LED label, the pixel coordinate respectively in three-dimensional tracking device middle left and right two width image; the coordinate of LED label in spatial digitizer coordinate system.

4d. is according to described projection relation and T ^twith the contextual definition two between Y for becoming the matrix of least square method matrix relationship with vectorial Y shape:

$A=\left(\begin{array}{c}(P_l^{11}-x_l^{i_{\mathrm{\λ}}}{P}_l^{31})x_S^{i_{\mathrm{\λ}}},(P_l^{11}-x_l^{i_{\mathrm{\λ}}}{P}_l^{31})x_S^{i_{\mathrm{\λ}}},(P_l^{11}-x_l^{i_{\mathrm{\λ}}}{P}_l^{31})x_S^{i_c},(P_l^{11}-x_l^{i_{\mathrm{\λ}}}{P}_l^{31}),(P_l^{12}-x_l^{i_{\mathrm{\λ}}}{P}_l^{32})x_S^{i_{\mathrm{\λ}}},(P_l^{12}-x_l^{i_{\mathrm{\λ}}}{P}_l^{32})x_S^{i_y},(P_l^{12}-x_l^{i_{\mathrm{\λ}}}{P}_l^{32})x_S^{i_c},(P_l^{12}-x_l^{i_{\mathrm{\λ}}}{P}_l^{32}),(P_l^{13}-x_l^{i_{\mathrm{\λ}}}{P}_l^{33})x_S^{i_{\mathrm{\λ}}},(P_l^{13}-x_l^{i_{\mathrm{\λ}}}{P}_l^{33})x_S^{i_y},(P_l^{13}-x_l^{i_{\mathrm{\λ}}}{P}_l^{33})x_S^{i_c}(P_l^{13}-x_l^{i_{\mathrm{\λ}}}{P}_l^{33})\\ (P_l^{21}-x_l^{i_{\mathrm{\λ}}}{P}_l^{31})x_S^{i_{\mathrm{\λ}}},(P_l^{21}-x_l^{i_{\mathrm{\λ}}}{P}_l^{31})x_S^{i_{\mathrm{\λ}}},(P_l^{21}-x_l^{i_{\mathrm{\λ}}}{P}_l^{31})x_S^{i_c},(P_l^{21}-x_l^{i_{\mathrm{\λ}}}{P}_l^{31}),(P_l^{22}-x_l^{i_{\mathrm{\λ}}}{P}_l^{32})x_S^{i_{\mathrm{\λ}}},(P_l^{22}-x_l^{i_{\mathrm{\λ}}}{P}_l^{32})x_S^{i_y},(P_l^{22}-x_l^{i_{\mathrm{\λ}}}{P}_l^{32})x_S^{i_c},(P_l^{22}-x_l^{i_{\mathrm{\λ}}}{P}_l^{32}),(P_l^{23}-x_l^{i_{\mathrm{\λ}}}{P}_l^{33})x_S^{i_{\mathrm{\λ}}},(P_l^{23}-x_l^{i_{\mathrm{\λ}}}{P}_l^{33})x_S^{i_y},(P_l^{23}-x_l^{i_{\mathrm{\λ}}}{P}_l^{33})x_S^{i_c},(P_l^{23}-x_l^{i_{\mathrm{\λ}}}{P}_l^{33})\\ (P_r^{11}-x_r^{i_{\mathrm{\λ}}}{P}_r^{31})x_S^{i_{\mathrm{\λ}}},(P_r^{11}-x_r^{i_{\mathrm{\λ}}}{P}_r^{31})x_S^{i_y},(P_r^{11}-x_r^{i_{\mathrm{\λ}}}{P}_r^{31})x_S^{i_c},(P_r^{11}-x_r^{i_{\mathrm{\λ}}}{P}_r^{31}),(P_r^{12}-x_r^{i_{\mathrm{\λ}}}{P}_r^{32})x_S^{i_{\mathrm{\λ}}},(P_r^{12}-x_r^{i_{\mathrm{\λ}}}{P}_r^{32})x_S^{i_y},(P_r^{12}-x_r^{i_{\mathrm{\λ}}}{P}_r^{32})x_S^{i_c},(P_r^{12},x_r^{i_{\mathrm{\λ}}}{P}_r^{32}),(P_r^{13}-x_r^{i_{\mathrm{\λ}}}{P}_r^{33})x_S^{i_{\mathrm{\λ}}},(P_r^{13}-x_r^{i_{\mathrm{\λ}}}{P}_r^{33})x_S^{i_y},(P_r^{13}-x_r^{i_{\mathrm{\λ}}}{P}_r^{33})x_S^{i_c},(P_r^{13}-x_r^{i_{\mathrm{\λ}}}{P}_r^{33})\\ (P_r^{21}-x_r^{i_{\mathrm{\λ}}}{P}_r^{31})x_S^{i_{\mathrm{\λ}}},(P_r^{21}-x_r^{i_{\mathrm{\λ}}}{P}_r^{31})x_S^{i_y},(P_r^{21}-x_r^{i_{\mathrm{\λ}}}{P}_r^{31})x_S^{i_c},(P_r^{21}-x_r^{i_{\mathrm{\λ}}}{P}_r^{31}),(P_r^{22}-x_r^{i_{\mathrm{\λ}}}{P}_r^{32})x_S^{i_{\mathrm{\λ}}},(P_r^{22}-x_r^{i_{\mathrm{\λ}}}{P}_r^{32})x_S^{i_y},(P_r^{22}-x_r^{i_{\mathrm{\λ}}}{P}_r^{32})x_S^{i_c},(P_r^{22}-x_r^{i_{\mathrm{\λ}}}{P}_r^{33})x_S^{i_{\mathrm{\λ}}},(P_r^{23}-x_r^{i_{\mathrm{\λ}}}{P}_r^{33})x_S^{i_y},(P_r^{23}-x_r^{i_{\mathrm{\λ}}}{P}_r^{33})x_S^{i_c},(P_r^{23}-x_r^{i_{\mathrm{\λ}}}{P}_r^{33})\end{array}\right)---\left(9\right)$

$b=\left(\begin{array}{c}(x_l^{i_x}{P}_l^{34}-P_l^{14})\\ (x_l^{i_y}{P}_l^{34}-P_l^{24)}\\ (x_r^{i_x}{P}_r^{34}-P_r^{14})\\ (x_r^{i_x}{P}_r^{34}-P_r^{24})\end{array}\right)---\left(10\right)$

And utilize formula (9) (10) that formula (5) is rewritten into least square method matrix relationship formula (11):

AY＝b (11)

4e uses least square method to minimize the attitude of formula (12) calculating spatial digitizer:

Y ^*＝argmin _Y(|AY-b| ²) (12)

4f. assesses the spatial digitizer attitudes vibration of two local measurements according to the attitude of the spatial digitizer calculated during double local measurement.

9th step, server 40, according to the result of the 8th step, splices the result of twice local measurement.

Tenth step, repeats step 5 to nine, until whole steel plate 10 three-dimensional measurement terminates, realizes the three-dimensional measurement to all local of steel plate 10 and splicing, obtain the complete three-D profile of large scale plate with curved surface 10.

In sum, technical scheme of the present invention, by spatial digitizer being arranged to track-movable, also distance object under test is comparatively near, comes flexibly, adopts method of structured light to enter the capable local measurement of piecemeal to large scale object easily, obtain high definition topography, simultaneously, because the light of LED label is enough bright, when the distance of three-dimensional tracking instrument and LED label increases, three-dimensional tracking instrument also can be quick, follow the tracks of the high-brightness LED label on spatial digitizer exactly, the center of LED label and the position in spatial digitizer coordinate system thereof accurately can be located with Deterministic service device, LED label relation between accurate acquisition two local measurements, and then the attitudes vibration of accurate evaluation spatial digitizer, realize the accurate splicing to local measurement result, therefore without the need to carrying out the loaded down with trivial details switching between multiple projector, just can facilitate, stable, accurately complete the three-dimensional measurement to the whole profile of large scale object.The local measurement of technical scheme of the present invention and the precision of image mosaic are all greatly improved.

Obviously, those skilled in the art can carry out various change and modification to invention and not depart from the spirit and scope of the present invention.Like this, if these amendments of the present invention and modification belong within the scope of the claims in the present invention and equivalent technologies thereof, then the present invention is also intended to comprise these change and modification.

Claims (9)

1., based on the three-dimensional geometry measuring method that LED label is followed the tracks of, it is characterized in that, comprising:

The first step, being arranged on by a set of spatial digitizer posting LED label can along on the support of the rail moving of object under test side, and by described support-moving one end to described track, described spatial digitizer mainly contains and is made up of two industrial cameras and a projector;

Second step, is fixedly installed three-dimensional tracking instrument, and the fixed position of three-dimensional tracking instrument makes can trace into described LED label all the time at measuring process neutral body tracker;

3rd step, start a local measurement, the projector of described spatial digitizer is to object under test surface projective structure light, two industrial cameras of described spatial digitizer take the object under test topography in the respective visual field, and described partial image data is transferred to server, meanwhile, three-dimensional tracking instrument is taken LED label image and is transferred to server;

4th step, described server generates the 3-D view of a local of described object under test according to the partial image data of described two industrial cameras, in described LED label image, accurately locate LED tag hub point position simultaneously, complete the measurement of a local of described object under test;

5th step, promotes support-moving along track and changes to reposition to promote spatial digitizer, repeats third and fourth step, completes the measurement of the next local of object under test;

6th step, described server adopts KLT algorithm, the LED label relation between the LED label image setting up two local measurements;

7th step, described server obtains the LED label position in the spatial digitizer coordinate system of each local measurement according to the LED label image of each local measurement;

8th step, described server with the LED label position of two local measurements, obtains the LED label position change of described two local measurements, to assess the spatial digitizer attitudes vibration of described two local measurements according to the LED label relation of two local measurements;

9th step, described server, according to the result of the 8th step, splices the 3-D view of described two local measurements;

Tenth step, repeats step 5 to nine, until last local measurement of described object under test terminates, to splice the 3-D view on the whole surface of described object under test.

2. three-dimensional geometry measuring method as claimed in claim 1, it is characterized in that, described spatial digitizer mainly contains and is made up of two industrial cameras and a projector.

3. three-dimensional geometry measuring method as claimed in claim 1, it is characterized in that, the luminous power of described LED label is more than or equal to 3W.

4. three-dimensional geometry measuring method as claimed in claim 1, it is characterized in that, described three-dimensional tracking instrument forms primarily of two industrial cameras.

5. three-dimensional geometry measuring method as claimed in claim 4, is characterized in that, the camera resolution of described spatial digitizer and described three-dimensional tracking instrument all more than 1440 × 1080, frame per second all at more than 10fps, and respectively by kilomega network connection server; The projector of described spatial digitizer is by USB line connection server.

6. three-dimensional geometry measuring method as claimed in claim 1, is characterized in that, in described 4th step, in described LED label image, accurately the step of LED tag hub point position, location comprises:

2a: adopt normalization template matching method, find the LED label image block in described LED label image;

2b: amplify described LED label image block and obtain enlarged image Pm, and in 3-D view corresponding to Pm, LED label Lm is rendered as octagonal mountain shape;

2c: adopt Gaussian function to the smoothing process of the 3-D view that Pm is corresponding, obtain a smoothed image, and determine the summit Pt of Lm image according to smoothed image summit;

2d: face based on 3-D view PV: one bottom surface Pb according to described level and smooth Computer image genration with five virtual planes, four cross facets all intersect at same intersection L perpendicular to Pb, and the angle between adjacent two crossing plane is 45 degree;

2e: superimposed images Lm and PV, and the bottom of Lm and PV overlaps and allows L by the summit Pt of Lm image;

2f: allow four cross facets of PV take L as axle, entirety rotates predetermined angle in the direction of the clock successively until rotate a circle, and records the intersection cross-sectional area of each postrotational PV and Lm;

2g: the current location of L and maximum intersection cross-sectional area part are recorded in queue Qs;

2h: 24 neighbor pixels representing Pt with Npt, mobile PV, allow L by each pixel in Npt, repeats step 2f ~ 2h;

2i: in Qs, selects the element with maximum intersection cross-sectional area, selects the position of the L that described element is corresponding as true center Ca;

2j: reduce the multiplying power of Pm to described LED label image block, and obtain the LED tag hub point position in described LED label image according to Ca.

7. three-dimensional geometry measuring method as claimed in claim 1, it is characterized in that, in described 7th step, described server comprises the step that the LED label position in the spatial digitizer coordinate system of each local measurement is demarcated:

3a: a gridiron pattern scaling board is placed in the visual field of three-dimensional tracking instrument and spatial digitizer, three-dimensional tracking instrument can also trace into the LED label on spatial digitizer simultaneously;

3b: obtain the transformational relation between three-dimensional tracking instrument coordinate system and spatial digitizer coordinate system by described gridiron pattern scaling board;

3c: the position of LED label in three-dimensional tracking instrument coordinate system being obtained each local measurement by the LED label image of three-dimensional tracking instrument shooting during each local measurement, and the position of LED label in three-dimensional tracking instrument coordinate system is converted to the position of LED label in spatial digitizer coordinate system by described transformational relation.

8. three-dimensional geometry measuring method as claimed in claim 1, is characterized in that, in described 8th step, according to the change of the described LED label position of described two local measurements, assess the spatial digitizer attitudes vibration of described two local measurements, comprising:

4a: by the position T of spatial digitizer during certain local measurement ^tbe expressed as vectorial Y, wherein,

$T^t=\left(\begin{array}{cccc}{T}_{11}^t& T_{12}^t& T_{13}^t& T_{14}^t\\ T_{21}^t& T_{22}^t& T_{23}^t& T_{24}^t\\ T_{31}^t& T_{32}^t& T_{33}^t& T_{34}^t\\ 0& 0& 0& 1\end{array}\right),$ $Y=(T_{11}^T,T_{12}^T,T_{13}^T,T_{14}^T,T_{21}^T,T_{22}^T,T_{23}^T,T_{24}^T,T_{31}^T,T_{32}^T,T_{33}^T,T_{34}^T);$

4b: described in definition, during certain local measurement, the left and right projection matrix of three-dimensional tracking instrument is respectively:

$P_l=\left(\begin{array}{cccc}{P_l}^{11}& {P_l}^{12}& {P_l}^{13}& {P_l}^{14}\\ {P_l}^{21}& {P_l}^{22}& {P_l}^{23}& {P_l}^{24}\\ {P_l}^{31}& {P_l}^{32}& {P_l}^{33}& {P_l}^{34}\end{array}\right);$ $P_r=\left(\begin{array}{cccc}{P_r}^{11}& {P_r}^{12}& {P_r}^{13}& {P_r}^{14}\\ {P_r}^{21}& {P_r}^{22}& {P_r}^{23}& {P_r}^{24}\\ {P_r}^{31}& {P_r}^{32}& {P_r}^{33}& {P_r}^{34}\end{array}\right);$

4c: according to P _l, P _rand T ^tby the coordinate X of LED label in spatial digitizer coordinate system during certain local measurement described _s ⁱproject in three-dimensional tracking instrument, set up projection relation:

$\left\{\begin{array}{c}{P}_l{T}^t{X}_S^i=x_l^i\\ P_r{T}^t{X}_S^i=x_r^i\end{array}\right.,i\∈1,...,N;$

Wherein, $x_l^i={\left(\begin{array}{cc}{x}_i^{i_x}& x_l^{i_y}\end{array}\right)}^T$ With $x_r^i={\left(\begin{array}{cc}{x}_r^{i_x}& x_r^{i_y}\end{array}\right)}^T$ The pixel coordinate of LED label respectively in three-dimensional tracking instrument middle left and right two width image; $X_S^i={\left(\begin{array}{ccc}{X}_S^{i_x}& X_S^{i_y}& X_S^{i_z}\end{array}\right)}^T$ The coordinate of LED label in spatial digitizer coordinate system;

4d: according to described projection relation and T ^tand the relation between Y obtain vectorial Y with p _l, P _rbetween least square method matrix relationship AY=b, wherein,

$b=\left(\begin{array}{c}\begin{array}{c}(x_l^{i_x}{P_l}^{34}-{P_l}^{14})\\ (x_l^{i_y}{P_l}^{34}-{P_l}^{24})\\ (x_r^{i_x}{P_r}^{34}-{P_r}^{14})\\ (x_r^{i_x}{P_r}^{34}-{P_r}^{24})\end{array}\end{array}\right);$

4e: the attitude calculating the spatial digitizer of certain local measurement described according to described least square method matrix relationship is:

Y ^*＝arg min _Y(|AY-b| ²)；

4f: the spatial digitizer attitudes vibration assessing two local measurements according to the attitude of the spatial digitizer calculated during double local measurement.

9. the three-dimensional geometry measuring system of following the tracks of based on LED label of the three-dimensional geometry measuring method based on the tracking of LED label according to any one of an application rights requirement 1 to 8, it is characterized in that, comprise a station server, a set of three-dimensional tracking instrument and a set of spatial digitizer posting LED label; Wherein, described spatial digitizer is arranged on can along on the support of the rail moving of object under test side, mainly contains and can be made up of to the projector of described object under test surface projective structure light two industrial cameras can taking the topography of object under test and one; Three-dimensional tracking instrument is fixedly installed and its fixed position makes can photograph described LED label image all the time at measuring process neutral body tracker; Two industrial cameras of three-dimensional tracking instrument and spatial digitizer are respectively by kilomega network connection server, and projector is by USB line connection server; Described server comprises:

According to the partial 3 d image composing unit of topography's synthesis partial 3 d image of two industrial camera shootings;

The KLT algorithm unit of the LED label relation between the LED label image that the three-dimensional tracking instrument setting up two local measurements is taken;

The LED obtaining the LED label position in the spatial digitizer coordinate system of each local measurement according to the LED label image of each local measurement demarcates unit;

The attitude assessment unit of the change obtaining the described LED label position of described two local measurements and the spatial digitizer attitudes vibration assessing described two local measurements;

Splice the image mosaic unit of the 3-D view of continuous two local measurements.