CN201576308U - Registration system for complex curved surface - Google Patents

Registration system for complex curved surface Download PDF

Info

Publication number
CN201576308U
CN201576308U CN2009202101641U CN200920210164U CN201576308U CN 201576308 U CN201576308 U CN 201576308U CN 2009202101641 U CN2009202101641 U CN 2009202101641U CN 200920210164 U CN200920210164 U CN 200920210164U CN 201576308 U CN201576308 U CN 201576308U
Authority
CN
China
Prior art keywords
registration
curved surface
solver
model
curved
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Expired - Lifetime
Application number
CN2009202101641U
Other languages
Chinese (zh)
Inventor
郭慧
林大钧
潘家祯
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
East China University of Science and Technology
Original Assignee
East China University of Science and Technology
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by East China University of Science and Technology filed Critical East China University of Science and Technology
Priority to CN2009202101641U priority Critical patent/CN201576308U/en
Application granted granted Critical
Publication of CN201576308U publication Critical patent/CN201576308U/en
Anticipated expiration legal-status Critical
Expired - Lifetime legal-status Critical Current

Links

Images

Landscapes

  • Length Measuring Devices With Unspecified Measuring Means (AREA)

Abstract

The utility model relates to a registration system for a complex curved surface, which comprises a curved surface CAD modeling device, a three-coordinate measuring machine and a registration solver. The curved surface CAD modeling device which includes geometric information of a curved surface is connected with a port at one edge of the registration solver, and the three-coordinate measuring machine is connected with the port at the other edge of the registration solver. Compared with the prior art, the registration system has the following characteristics: a theoretical model is accurate, the registration stability is high, the registration accuracy is good, the application range is wide, and the multi-dimensional variable registration can be realized.

Description

A kind of complex-curved registration arrangement
Technical field
The utility model relates to Machine Design and detection technique field, relates in particular to a kind of complex-curved registration arrangement.
Background technology
Complex-curved registration technology is widely used in the fields such as shape error detection, multi-viewpoint cloud reorientation, reverse-engineering, virtual reality, and it is the precondition of implementation model assessment, error analysis and data amalgamation.Because cad model designs in CAD design software coordinate system, measurement is to carry out in the coordinate system of measuring equipment, complex-curved measurement coordinate system objectively is being separate with design coordinate system (CAD coordinate system), measurement data and cad model do not have clear and definite size corresponding relation, add the complex-curved tangible reference characteristic that do not have, complex-curved mathematical model is very complicated, be difficult to find size corresponding relation accurately between complex-curved measuring basis and design basis, therefore complex-curved registration is key to the issue and difficult point always.If the two disunity will inevitably influence the result of calculation of follow-up error-detecting, model evaluation etc.
More common method for registering is least square method and iterative closest point method at present, and these two kinds of methods all come with some shortcomings.
The ultimate principle of least square method is to be target with the residual sum of squares (RSS) minimum between the two, have and calculate advantage simple, that be easy to realize, but it is not high enough by least square method result calculated precision, often than the big 1.8%-30% of precise results, thereby be not suitable for the high complex-curved registration of accuracy requirement.
The iterative closest point method at first must be found out two corresponding subclass between point set and point set, obtain the transformation matrix of two point sets then, its algorithm is simple, precision is higher, its less than one is the subclass that point set is the another one point set that two match points of requirement are concentrated, the 2nd, the relative initial position of two point sets is had relatively high expectations, the relative initial position of two point sets can not differ too big, when two groups of point set initial positions differ big, easily be absorbed in local minimum, and the convergence direction may be uncertain.
Summary of the invention
The purpose of this utility model is exactly to provide a kind of stability high, good, the widely used complex-curved registration arrangement of degree of accuracy for the defective that overcomes above-mentioned prior art existence.
The purpose of this utility model can be achieved through the following technical solutions:
A kind of complex-curved registration arrangement, it is characterized in that, this system comprises curved surface CAD model building device, three coordinate measuring machine and registration solver, described curved surface CAD model building device comprises the geological information of curved surface, be connected with the port on one side of registration solver, described three coordinate measuring machine is connected with the port of the another side of registration solver.
Described curved surface CAD model building device is the computing machine that the CAD design software is installed.
Compared with prior art, the utlity model has following advantage:
(1) theoretical model is accurate: cad model is the foundation of optimal comparison, and nurbs surface has extensive versatility, and setting up its theoretical model according to the CAD design information of curved surface has crucial effects to error evaluation;
(2) stability of registration: utilize robustness, the concurrency of genetic algorithm and characteristics with ability of searching optimum, complexity is carried out the thick registration of curved surface, can seek appropriate initial registration conversion fast, overcome traditional iterative closest point method to the demanding shortcoming of initial value, can be from the arbitrary initial relative position, the convergence that assurance is found the solution has improved counting yield, can accelerate to obtain desirable registration results;
(3) accuracy of registration: behind the thick registration again the using iterative closest point algorithms carry out the smart registration of secondary.Utilize the higher characteristics of iterative closest point law technology precision, overcome genetic algorithm method later stage long deficiency consuming time, improved accuracy of registration;
(4) realize the multidimensional variable registration: can obtain the result that simple iterative closest point method can't obtain, the multivariate registration in the three-dimensional data registration is had stronger robustness;
(5) popularity of Ying Yonging: can be applicable to NURBS represent complex-curved arbitrarily, the nurbs surface expression way not only has character such as partial controllable, standardization, differentiability, also the accurate expression for the parsing shape (as quadric surface, the surface of revolution etc.) of standard curved surface and free form surface provides a public mathematical form with design, in CAD/CAM and field of Computer Graphics, nurbs surface is widely used.
Description of drawings
Fig. 1 is the synoptic diagram of this system;
Fig. 2 is model data and the measurement data before recessed registration of turbine blade;
Fig. 3 is the thick registration results of genetic algorithm iteration 50 times;
Fig. 4 is the result of the smart registration of iterative closest point;
Fig. 5 is simple ICP registration results.
Embodiment
Below in conjunction with the drawings and specific embodiments the utility model is elaborated.
Embodiment
A kind of complex-curved registration arrangement, its structure as shown in Figure 1, this system comprises curved surface CAD model building device, three coordinate measuring machine and registration solver, curved surface CAD model building device is the computing machine that the CAD design software is installed, the geological information that comprises curved surface adopts the NURBS function to make up curved surface NURBS model, is connected with the port on one side of registration solver, three-dimensional coordinates measurement machine testing curved surface prototype in kind is also handled and is obtained the measure dot number certificate, is connected with the port of registration solver another side.Data after the registration solver adopts the genetic algorithm comparison, calculates and revises curved surface NURBS model and handles.
Curved surface CAD modeling comprises the geological information of curved surface, utilize IGES file or DXF file to extract control vertex, node and weight factor information, make up curved surface NURBS model, the curved surface form that adopts the NURBS function representation to design a model, and nurbs surface dispersed along parametric direction, obtain an accurate goal directed point cloud, three-dimensional coordinates measurement machine testing curved surface prototype in kind, obtain the measure dot number certificate, the treated measurement point cloud that obtains, the registration solver adopts genetic algorithm that goal directed is put cloud and the measurement point cloud carries out thick registration, the registration transformation that obtains, measurement data is carried out initial transformation, obtain the measurement point data set behind the thick registration, the using iterative closest point algorithms is carried out smart registration and is calculated again, seek behind the thick registration measurement data and design a model nearest point set, adopt singular value decomposition method to calculate optimal transformation, till satisfying stopping criterion for iteration, obtain final registration results.
Adopt this system that the turbine blade concave surface is carried out registration, registration front end model data and measurement data figure are as shown in Figure 2, Fig. 2 has expressed curved design model before the registration and the relative position turbine blade concave surface cad model of measurement data is a smooth surface, to the forward face 3 d measurement data of gathering 900 curved surfaces in kind, the some cloud is depicted as measurement data in kind.
Utilization genetic algorithm method is carried out thick registration and is calculated, and the registration result calculated sees Table 1, and wherein iterations is that 50 o'clock registration effect is seen shown in Figure 3.
With the result of the genetic algorithm method iteration 50 times initial position as smart registration, the using iterative closest point algorithms is carried out the smart registration of secondary, and only the iteration optimum solution that just obtains registration 10 times sees Table 1, and the effect that realizes the secondary registration with this result as shown in Figure 4.
The registration results of table 1 spoon of blade cad model
Figure G2009202101641D00041
And use merely that the ICP algorithm carries out a registration see Table 1, no longer continue to seek optimum solution after its result reaches 0.8867.Its registration results as shown in Figure 5, measurement point cloud and CAD curved surface direction are reverse, illustrate that ICP convergence of algorithm direction is unsettled, can not get optimum solution sometimes when measurement data and theoretical model initial position differ far away.
Comparison sheet 1 result as can be known, when measurement point and CAD theoretical model relative position differ big, only use iterative closest point algorithms, may be absorbed in local minimum, can't obtain best alignment, at this moment, adopt the genetic algorithm method to carry out thick registration, it is very effective adopting iterative closest point algorithms to carry out smart registration then, and the global optimizing performance of genetic algorithm method can guarantee that the secondary registration strategies obtains globally optimal solution.Relatively the result of the thick registration of genetic algorithm method 50 times, 100 times as can be known, as approximate solution, the two is more approaching.
In view of the genetic algorithm method in the slower shortcoming of computing late convergence, from the consideration of saving time, after the genetic algorithm method is moved thick registration of 50 generations, in conjunction with iterative closest point algorithms refinement computing, can guarantee to obtain optimum solution again, and higher efficient is arranged.
Fig. 3 shows, adopts the genetic algorithm method to carry out thick registration, and its convergence direction is correct, and as shown in Figure 4, after the calculating of secondary registration, measurement point and cad model have been realized best alignment.

Claims (2)

1. complex-curved registration arrangement, it is characterized in that, this system comprises curved surface CAD model building device, three coordinate measuring machine and registration solver, described curved surface CAD model building device is connected with a port of registration solver, and described three coordinate measuring machine is connected with the another port of registration solver.
2. a kind of complex-curved registration arrangement according to claim 1 is characterized in that, described curved surface CAD model building device is the computing machine that the CAD modeling software is installed.
CN2009202101641U 2009-09-25 2009-09-25 Registration system for complex curved surface Expired - Lifetime CN201576308U (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN2009202101641U CN201576308U (en) 2009-09-25 2009-09-25 Registration system for complex curved surface

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN2009202101641U CN201576308U (en) 2009-09-25 2009-09-25 Registration system for complex curved surface

Publications (1)

Publication Number Publication Date
CN201576308U true CN201576308U (en) 2010-09-08

Family

ID=42696184

Family Applications (1)

Application Number Title Priority Date Filing Date
CN2009202101641U Expired - Lifetime CN201576308U (en) 2009-09-25 2009-09-25 Registration system for complex curved surface

Country Status (1)

Country Link
CN (1) CN201576308U (en)

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104316016A (en) * 2014-11-03 2015-01-28 沈阳黎明航空发动机(集团)有限责任公司 Longitudinal measurement method for blisk complex curved surface blade
CN107749082A (en) * 2017-10-31 2018-03-02 黑龙江省科学院自动化研究所 A kind of complex-curved registration arrangement
CN112066923A (en) * 2020-08-14 2020-12-11 苏州环球科技股份有限公司 Virtual measurement method for position degree
US20230063166A1 (en) * 2021-08-31 2023-03-02 Mloptic Corp Method for measuring the surface shape of a toric mirror

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104316016A (en) * 2014-11-03 2015-01-28 沈阳黎明航空发动机(集团)有限责任公司 Longitudinal measurement method for blisk complex curved surface blade
CN104316016B (en) * 2014-11-03 2017-08-04 沈阳黎明航空发动机(集团)有限责任公司 A kind of longitudinal measuring method of the complex-curved blade of blisk
CN107749082A (en) * 2017-10-31 2018-03-02 黑龙江省科学院自动化研究所 A kind of complex-curved registration arrangement
CN112066923A (en) * 2020-08-14 2020-12-11 苏州环球科技股份有限公司 Virtual measurement method for position degree
US20230063166A1 (en) * 2021-08-31 2023-03-02 Mloptic Corp Method for measuring the surface shape of a toric mirror
US11934743B2 (en) * 2021-08-31 2024-03-19 Mloptic Corp. Method for measuring the surface shape of a toric mirror

Similar Documents

Publication Publication Date Title
CN101692257B (en) Method for registering complex curved surface
Li et al. A modified ICP algorithm based on dynamic adjustment factor for registration of point cloud and CAD model
Ji et al. An improved method for registration of point cloud
Ge et al. Surface-based matching of 3D point clouds with variable coordinates in source and target system
CN102622747B (en) Camera parameter optimization method for vision measurement
CN103295232B (en) Based on the SAR image registration method in straight line and region
Li et al. A stable direct solution of perspective-three-point problem
CN102169579A (en) Rapid and accurate registration method of dense point cloud model
CN107044852B (en) Total station survey method under out-of-flatness state
Yuan et al. 3D point cloud matching based on principal component analysis and iterative closest point algorithm
CN103106632B (en) A kind of fusion method of the different accuracy three dimensional point cloud based on average drifting
CN101882150B (en) Three-dimensional model comparison and search method based on nuclear density estimation
CN201576308U (en) Registration system for complex curved surface
CN110595479B (en) SLAM track evaluation method based on ICP algorithm
CN108332685B (en) A kind of coding structural light three-dimensional measurement method
CN110599582A (en) Quick algorithm of quasi-rectangular object bounding box
CN103292769B (en) A kind of plane inclination error detection method based on Minimum Area
Yun et al. Registration of multiview point clouds for application to ship fabrication
CN103808286A (en) Total station-based steel structure three dimensional precision detection analysis method and application thereof
CN104318552A (en) Convex hull projection graph matching based model registration method
Ao et al. A repeatable and robust local reference frame for 3D surface matching
CN103942784A (en) Method for solving parameters in parabolic catadioptric camera through three unparallel straight lines in checkerboard
CN102968400A (en) Multi-view three-dimensional data registration method based on spatial line recognition and matching
CN109163708B (en) Optimization method, device and medium for unmanned aerial vehicle large-scale topographic map house line
CN103487791A (en) RCS conversion method based on scattering center matrix

Legal Events

Date Code Title Description
C14 Grant of patent or utility model
GR01 Patent grant
CX01 Expiry of patent term

Granted publication date: 20100908

CX01 Expiry of patent term