CN103955939B - 3 D scanning system midpoint cloud edge feature point method for registering - Google Patents
3 D scanning system midpoint cloud edge feature point method for registering Download PDFInfo
- Publication number
- CN103955939B CN103955939B CN201410208338.6A CN201410208338A CN103955939B CN 103955939 B CN103955939 B CN 103955939B CN 201410208338 A CN201410208338 A CN 201410208338A CN 103955939 B CN103955939 B CN 103955939B
- Authority
- CN
- China
- Prior art keywords
- point
- cloud
- matrix
- point set
- registration
- 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 - Fee Related
Links
Landscapes
- Image Analysis (AREA)
Abstract
The invention discloses a kind of 3 D scanning system midpoint cloud edge feature point method for registering, step is as follows：1) three-dimensional laser scanner obtains the spatial sampling point on different visual angles real-world object surface；2) the point cloud boundary characteristics point of the boundary detection method extraction different visual angles of point of use cloud centroidal distance feature；3) closest point (ICP) the algorithm registration point cloud of improved iteration is used according to the point cloud boundary characteristics of extraction point；4) registration accuracy is evaluated according to registration error standard, whether verification registration result reaches with alignment request.The present invention carries out edge feature point extraction by treating registration point cloud, the defects of each point searches corresponding points in point cloud data need to be traversed by avoiding traditional ICP algorithm, on the basis of registration accuracy is ensured, the efficiency of point cloud registering is significantly improved while effectively reducing algorithm complexity.
Description
Technical field
The invention belongs to the technical fields such as reverse-engineering, image procossing, and in particular to a kind of 3 D scanning system point cloud
Splicing edge feature point method for registering.
Background technology
The three-dimensional geometry that point cloud data obtains body surface discrete point generally by measuring instruments such as spatial digitizers is sat
Mark due to being limited by factors such as size, environment and the measuring appliances for measuring object, can only measure a side of object every time
Therefore face, to obtain the partial data information of testee, needs to be scanned object from multiple and different angles.It will be different
Point cloud data under visual angle is uniformly the registration of point cloud data to the process in the same coordinate system.Along with reconstructing three-dimensional model
In the extensive use of the various fields such as reverse-engineering, industrial detection, Medical Image Processing, historical relic's protection, three dimensional point cloud is matched
Quasi- technology also becomes popular and important research topic.
Existing cloud data registration technology is broadly divided into：Manual registration relies on instrument registration and autoregistration, wherein automatically
Registration technique obtains most commonly used application, which mainly includes two steps of rough registration and accuracy registration.Rough registration is
The premise of good initial value is provided for accuracy registration, accuracy registration is the key that then to determine point cloud registering error, associated specialist
Person has done a large amount of correlative studys in terms of accuracy registration.It is at present the closest point of iteration with most commonly used accuracy registration algorithm
Algorithm (iterative close point, ICP).ICP algorithm proposes respectively by Besl et al. and Chen et al. earliest, but its
Shortcoming mainly has：The quality of initial point cloud directly influences splicing precision, and iteration takes, restrains relatively slow, registration is easily done by noise
It disturbs, be easily absorbed in local optimum etc..For obtain accuracy registration as a result, domestic and foreign scholars classical ICP algorithm has been done numerous studies and
It improves.Such as in terms of Irregular Boundary Surface point cloud registering, it is registrated using rough registration and the improved ICP algorithm of rigidity characteristic, it is real
The accuracy registration of Irregular Boundary Surface point cloud is showed.Although the convergence speed of the algorithm and joining quality are obtained for certain raising,
But influence of the noise to registration accuracy is not accounted for, robustness is poor in some cases.As combined based on curvature feature
The ICP innovatory algorithms of point can effectively be registrated unknown point cloud, have a clear superiority on Quasi velosity.It but must be first
Begin under the guarantee of registration, registration can just be made to be unlikely to tend to the opposite way round.Otherwise, when the rotary shifted of two panels point cloud and translation are wrong
When bit comparison is big, only registration can be caused to be absorbed in local optimum using only improved ICP accuracy registrations.Such as assign corresponding points weight and
The Revised ICP algorithm of M-estimation is introduced, accuracy registration is carried out with Revised ICP algorithm on the basis of rough registration.The innovatory algorithm
Solve the efficiency bottle neck of classical ICP algorithm, effectively eliminate influence of the abnormal point to algorithm, improve algorithm accuracy and
Reliability, but need to be further improved on proximity pair search efficiency.
Currently, closest point (ICP) method of existing iteration is：In the point cloud number that different visual angles are obtained by spatial digitizer
According to rear, the closest approach that two points are concentrated is found every time with ICP algorithm, and make the quadratic sum of its Euclidean distance minimum, so as to count
Calculate the rigid body translation between point cloud.The rigid body translation is applied to point cloud and obtains new target point cloud, if object function error is received
It holds back in given threshold value, terminates iteration, otherwise continue to search for closest approach.
As shown in Figure 1, Fig. 1, which is used, improves iterative closest point approach progress cloud data registration, however the point cloud registering side
Method has the following disadvantages and defect：
(1) algorithm complexity of algorithm extraction point cloud boundary characteristics point is higher and without improved search strategy, causes most
Near point is not high to search efficiency, so as to influence to be registrated efficiency.
(2) algorithm must can just make registration be unlikely to tend to the opposite way round under the guarantee of initial registration.Otherwise, when
When the rotary shifted and translation dislocation of two panels point cloud is bigger, only registration can be caused to be absorbed in using only improved ICP accuracy registrations
Local optimum.
Therefore, to obtain quickly and accurately registration result, boundary characteristic is obtained with new point cloud boundary extracting method
Point is searched for by K-D trees in edge feature point and accelerates to search proximity pair, can accelerated search speed, further improve and match
Quasi- efficiency.
Invention content
For above-mentioned deficiency in the prior art, the 3-D scanning system with Quasi velosity is improved the present invention provides a kind of
System midpoint cloud edge feature point method for registering.
In order to solve the above-mentioned technical problem, present invention employs following technical solutions：
3 D scanning system midpoint cloud edge feature point method for registering, this method comprises the following steps：
(1), it scans the profile of mock-up with different view with spatial digitizer, obtains multi-view angle three-dimensional sampling number
According to；
(2), for two pieces of three-dimensional sample point datas of mock-up same under different visual angles, by with a cloud distance of centre of gravity
Boundary detection method from feature extracts the edge feature point of two pieces of three-dimensional sample point datas respectively；
(3), in two block boundary characteristic points of extraction, it is with reference to point set to fix one of, and another piece is target point set,
Corresponding proximity pair is obtained using K-D tree methods acceleration search, and is calculated according to unit Quaternion method so that corresponding proximity pair
The rigid body translation of average distance minimum, the rigid body translation are spin matrix R and translation matrix T；
(4), coordinate transform is carried out to target point set using spin matrix R and translation matrix T, obtains new target point set,
New target point set is judged with whether the distance with reference to point set is less than given threshold value according to stopping criterion for iteration to realize a cloud
Registration；
(5), it can quickly be carried out a little come Revised ICP algorithm based on the boundary extraction method of cloud centroidal distance feature with this
Cloud is registrated, and obtains accuracy registration effect.
As a preferred embodiment of the present invention, the boundary detection method of described cloud centroidal distance feature, specific steps
It is as follows：
The numerical value that point cloud elevation information is represented on Z axis is converted into matter of the gray value between 0-255 as point cloud first
Attribute is measured, is denoted as H；Current point P (X are found secondly by k nearest neighbor searchi,Yi,Zi) K arest neighbors, K=16；Finally calculate K
The barycentric coodinates (X, Y, Z) of the point group of a arest neighbors composition, threshold value δ is obtained by maximum kind differences method；Point P is calculated to point group weight
The Euclidean distance of the heart, as the Euclidean distance of fruit dot P to point group center of gravity is more than threshold value δ, then it is assumed that point P is marginal point；Conversely, point P
It is not marginal point；
Barycentric coodinates (X, Y, Z) formula of the point group of wherein K arest neighbors composition is as follows：
Z=Hi
The Euclidean distance formula of point P to point group center of gravity is as follows：
As another preferred embodiment of the present invention, the unit Quaternion method is as follows：
Assuming that target point set X={ xi|xi∈R3, i=1,2 ... m }, with reference to point set Y={ yj|yj∈R3, j=1,2,
... m }, for target point set X with being corresponded with reference to the point in point set Y, m is point cloud data amount；
1) target point set X and the center of gravity with reference to point set Y are asked respectively；
2) covariance matrix is constructed；
Wherein：(yi-μY)TFor matrix (yi-μY) transposition；
3) 4 × 4 symmetrical matrixes are constructed according to covariance matrix；
Wherein：tr∑X,YIt is matrix ∑X,YMark, if Ai,j=(∑X,Y-∑X,Y T)i,j, Δ=[A23,A21,A12]T, I3It is 3
× 3 unit matrixs, Ai,jFor (∑X,Y-∑X,Y T)i,jThe matrix of construction；
4) 4 × 4 symmetrical matrix Q (∑s are calculatedX,Y) characteristic value and feature vector, the corresponding feature vector of maximum eigenvalue
As unit quaternion [q0,q1,q2,q3]T；Rotating vector is qR=[q0,q1,q2,q3]T；
5) spin matrix is calculated
6) translation matrix is calculated
T(qR)=[q4,q5,q6]T=μX-R(qR)μY。
As another preferred embodiment of the present invention, the iteration ends Rule of judgment is as follows：
If new transformation point set is less than given threshold value λ with the European average distance with reference to point set, stop iteration, it is on the contrary
New point set is then continued into iteration as initial value, until meeting the requirement of object function；Wherein object function is:
Wherein:N be point cloud data amount, miFor reference point clouds to be registered, R is spin matrix, and T is translation matrix, SiFor mesh
Punctuate cloud, i=1,2...n.
The beneficial effects of the invention are as follows：It needs to traverse each point for ICP algorithm to calculate corresponding points, algorithm calculation amount
Larger, the shortcomings that efficiency is very low, the present invention proposes a kind of Revised ICP algorithm of the edge detection based on gravity center characteristics for point
The accuracy registration of cloud data, this method not only solves the bottleneck on ICP algorithm search strategy, and is ensureing registration accuracy
On the basis of effectively increase registration efficiency；Meanwhile the simplification rate based on border detection point cloud data reaches 2.2431%, compared to
The registration efficiency of congenic method improves more than 25.8%, shows that this method is more suitable for the larger point cloud data of data volume and matches
It is accurate.
Description of the drawings
Fig. 1 is the schematic diagram of cloud data registration of the prior art；
Fig. 2 is the schematic diagram of existing boundary detection method and the embodiment one of boundary detection method of the present invention comparison；
The embodiment two that Fig. 3 is classical ICP algorithm, existing improvement ICP methods are compared with present invention improvement ICP methods is shown
It is intended to；
Fig. 4 is the implementation that existing improvement ICP methods are directed to the comparison of different visual angles point cloud registering with present invention improvement ICP methods
The schematic diagram of example three；
Fig. 5 is that existing ICP algorithm is registrated the schematic diagram for taking comparison with the method for the present invention.
Specific embodiment
The present invention is described in further detail with reference to the accompanying drawings and detailed description.
3 D scanning system midpoint cloud edge feature point method for registering, this method comprises the following steps：
(1), it scans the profile of mock-up with different view with spatial digitizer, obtains multi-view angle three-dimensional sampling number
According to；
(2), for two pieces of three-dimensional sample point datas of mock-up same under different visual angles, by with a cloud distance of centre of gravity
Boundary detection method from feature extracts the edge feature point of two pieces of three-dimensional sample point datas respectively；
(3), in two block boundary characteristic points of extraction, it is with reference to point set to fix one of, and another piece is target point set,
Corresponding proximity pair is obtained using K-D tree methods acceleration search, and is calculated according to unit Quaternion method so that corresponding proximity pair
The rigid body translation of average distance minimum, as spin matrix R and translation matrix T；
(4), coordinate transform is carried out to target point set using spin matrix R and translation matrix T, obtains new target point set,
New target point set is judged with whether the distance with reference to point set is less than given threshold value according to stopping criterion for iteration to realize a cloud
Registration；
(5), it can quickly be carried out a little come Revised ICP algorithm based on the boundary extraction method of cloud centroidal distance feature with this
Cloud is registrated, and obtains accuracy registration effect.
Based on a boundary detection method for cloud centroidal distance feature, step is as follows：
1) numerical value that point cloud elevation information is represented on Z axis is converted into quality of the gray value between 0-255 as point cloud
Attribute is denoted as H；
2) current point P (X are found by k nearest neighbor searchi,Yi,Zi) K arest neighbors, K=16；
3) barycentric coodinates (X, Y, Z) of the point group of K arest neighbors composition are calculated, formula is as follows：
Z=Hi
Threshold value δ is obtained by Otsu (maximum kind differences method)；
4) point P is calculated to the Euclidean distance of point group center of gravity, and formula is as follows：
5) as the Euclidean distance of fruit dot P to point group center of gravity is more than threshold value δ, then it is assumed that point P is marginal point；Conversely, point P is not
Marginal point.
Rigid body translation is carried out using unit Quaternion method, is as follows：
Assuming that target point set X={ xi|xi∈R3, i=1,2 ... m }, with reference to point set Y={ yj|yj∈R3, j=1,2,
... m }, for target point set X with being corresponded with reference to the point in point set Y, m is point cloud data amount；
1) target point set X and the center of gravity with reference to point set Y are asked respectively
2) covariance matrix is constructed；
Wherein：(yi-μY)TFor matrix (yi-μY) transposition；
3) 4 × 4 symmetrical matrixes are constructed according to covariance matrix；
Wherein tr ∑sX,YIt is matrix ∑X,YMark, if Ai,j=(∑X,Y-∑X,Y T)i,j, Δ=[A23,A21,A12]T, I3It is 3
× 3 unit matrixs, Ai,jFor (∑X,Y-∑X,Y T)i,jThe matrix of construction；
4) 4 × 4 symmetrical matrix Q (∑s are calculatedX,Y) characteristic value and feature vector, the corresponding feature vector of maximum feature is
For unit quaternary number [q0,q1,q2,q3]T；Rotating vector is qR=[q0,q1,q2,q3]T；
5) spin matrix is calculated
6) translation matrix is calculated
T(qR)=[q4,q5,q6]T=μX-R(qR)μY。
Iteration ends Rule of judgment is as follows：
If new transformation point set is less than given threshold value λ with the European average distance with reference to point set, stop iteration, it is on the contrary
New point set is then continued into iteration as initial value, until meeting the requirement of object function；Wherein object function is:
Wherein:N be point cloud data amount, miFor reference point clouds to be registered, R is spin matrix, and T is translation matrix, SiFor mesh
Punctuate cloud, i=1,2...n.
Present invention employs new point cloud boundary detection method Revised ICP algorithms.Based on a side for cloud centroidal distance feature
Boundary is detected as shown in Fig. 2, point cloud boundary extraction comparison is as shown in table 1 below；
Table 1：
By as shown in table 2 below to being compared after Fig. 3 progress accuracy registration algorithms；
Table 2：
| Method for registering | It is registrated time/ms | Registration error/mm | Iterations/time |
| Classical ICP algorithm | 112.1820 | 0.1544 | 137 |
| Revised ICP algorithm | 80.4597 | 0.1287 | 87 |
| This paper algorithms | 51.5570 | 0.0965 | 30 |
It is as shown in table 3 below by carrying out the comparison of different visual angles cloud data registration to Fig. 4；
Table 3：
From Fig. 3, Fig. 4 and Fig. 5 and table 2 and table 3 as can be seen that the present invention carries out cloud data registration with edge feature point
Method solves the bottleneck on ICP algorithm search strategy, and registration effect is effectively increased on the basis of registration accuracy is ensured
Rate.Meanwhile the simplification rate based on border detection point cloud data reaches 2.2431%, the registration efficiency compared to congenic method improves
More than 25.8%, show that the improved ICP algorithm of the present invention is more suitable for the larger cloud data registration of data volume.
Finally illustrate, the above embodiments are merely illustrative of the technical solutions of the present invention and it is unrestricted, although with reference to compared with
The present invention is described in detail in good embodiment, it will be understood by those of ordinary skill in the art that, it can be to the skill of the present invention
Art scheme is modified or replaced equivalently, and without departing from the objective and range of technical solution of the present invention, should all be covered at this
In the right of invention.
Claims (1)
1. 3 D scanning system midpoint cloud edge feature point method for registering, which is characterized in that this method includes following step
Suddenly：
(1), it scans the profile of mock-up with different view with spatial digitizer, obtains multi-view angle three-dimensional sample point data；
(2), for two pieces of three-dimensional sample point datas of mock-up same under different visual angles, by special with point cloud centroidal distance
The boundary detection method of sign extracts the edge feature point of two pieces of three-dimensional sample point datas respectively；Wherein, the boundary detection method
It is as follows：
The numerical value that point cloud elevation information is represented on Z axis is converted into quality category of the gray value between 0-255 as point cloud first
Property, it is denoted as H；Current point P (X are found secondly by k nearest neighbor searchi,Yi,Zi) K arest neighbors, K=16；Finally calculate K most
The barycentric coodinates (X, Y, Z) of the point group of neighbour's composition obtain threshold value δ by maximum kind differences method；Point P is calculated to point group center of gravity
Euclidean distance, as the Euclidean distance of fruit dot P to point group center of gravity is more than threshold value δ, then it is assumed that point P is marginal point；Conversely, point P is not
Marginal point；
Barycentric coodinates (X, Y, Z) formula of the point group of wherein K arest neighbors composition is as follows：
Z=Hi
The Euclidean distance formula of point P to point group center of gravity is as follows：
(3), it is fixed one of for reference to point set, another piece is target point set, is utilized in two block boundary characteristic points of extraction
K-D tree methods acceleration search obtains corresponding proximity pair, and is calculated according to unit quaternion method so that corresponding proximity pair is averaged
The minimum rigid body translation of distance, the rigid body translation are spin matrix R and translation matrix T；Wherein, the unit quaternion method, tool
Body step is as follows：
Assuming that target point set X={ xi|xi∈R3, i=1,2 ... m }, with reference to point set Y={ yj|yj∈R3, j=1,2 ... m },
For target point set X with being corresponded with reference to the point in point set Y, m is point cloud data amount；
1) target point set X and the center of gravity with reference to point set Y are asked respectively；
2) covariance matrix is constructed；
Wherein:(yi-μY)TFor matrix (yi-μY) transposition；
3) 4 × 4 symmetrical matrixes are constructed according to covariance matrix；
Wherein:tr∑X,YIt is matrix ∑X,YMark, if Ai,j=(∑X,Y-∑X,Y T)i,j, Δ=[A23,A21,A12]T, I3It is 3 × 3
Unit matrix, Ai,jFor (∑X,Y-∑X,Y T)i,jThe matrix of construction；
4) 4 × 4 symmetrical matrix Q (∑s are calculatedX,Y) characteristic value and feature vector, the corresponding feature vector of maximum eigenvalue is
Unit quaternion [q0,q1,q2,q3]T；Rotating vector is qR=[q0,q1,q2,q3]T；
5) spin matrix is calculated
6) translation matrix is calculated
T(qR)=[q4,q5,q6]T=μX-R(qR)μY；
(4), coordinate transform is carried out to target point set using spin matrix R and translation matrix T, obtains new target point set, according to
Stopping criterion for iteration judges new target point set with whether the distance with reference to point set is less than given threshold value to realize point cloud registering；
(5), a cloud can be quickly carried out based on the boundary extraction method of cloud centroidal distance feature come Revised ICP algorithm with this to match
Standard obtains accuracy registration effect；
The iteration ends Rule of judgment is as follows：
If new transformation point set is less than given threshold value λ with the European average distance with reference to point set, stop iteration, it is on the contrary then by
New point set continues iteration as initial value, until meeting the requirement of object function；Wherein object function is:
Wherein:N be point cloud data amount, miFor reference point clouds to be registered, R is spin matrix, and T is translation matrix, SiFor target point
Cloud, i=1,2...n.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410208338.6A CN103955939B (en) | 2014-05-16 | 2014-05-16 | 3 D scanning system midpoint cloud edge feature point method for registering |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410208338.6A CN103955939B (en) | 2014-05-16 | 2014-05-16 | 3 D scanning system midpoint cloud edge feature point method for registering |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103955939A CN103955939A (en) | 2014-07-30 |
| CN103955939B true CN103955939B (en) | 2018-06-19 |
Family
ID=51333208
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410208338.6A Expired - Fee Related CN103955939B (en) | 2014-05-16 | 2014-05-16 | 3 D scanning system midpoint cloud edge feature point method for registering |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103955939B (en) |
Families Citing this family (46)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104237380B (en) * | 2014-10-10 | 2017-01-18 | 北京理工大学 | Composite component manipulator scanning pose quaternion converting method |
| US9460517B2 (en) | 2014-10-22 | 2016-10-04 | Pointivo, Inc | Photogrammetric methods and devices related thereto |
| CN104392426B (en) * | 2014-10-23 | 2017-07-18 | 华中科技大学 | A kind of no marks point three-dimensional point cloud method for automatically split-jointing of self adaptation |
| CN104296659B (en) * | 2014-10-27 | 2017-01-18 | 苏州天准科技股份有限公司 | Method for correcting joint error through inclined standard face |
| CN104680530A (en) * | 2015-03-01 | 2015-06-03 | 江西科技学院 | ICP algorithm |
| CN104700451B (en) * | 2015-03-14 | 2017-05-17 | 西安电子科技大学 | Point cloud registering method based on iterative closest point algorithm |
| CN105719287B (en) * | 2016-01-19 | 2019-04-26 | 博众精工科技股份有限公司 | A kind of alignment method of irregular figure |
| CN105931234A (en) * | 2016-04-19 | 2016-09-07 | 东北林业大学 | Ground three-dimensional laser scanning point cloud and image fusion and registration method |
| CN106204603B (en) * | 2016-04-29 | 2018-06-29 | 北京仁光科技有限公司 | Three-dimensional camera solid matching method |
| CN107423773B (en) * | 2016-05-23 | 2020-02-14 | 北京师范大学 | Automatic registration method and device for three-dimensional skull |
| CN106204718B (en) * | 2016-06-28 | 2019-08-20 | 华南理工大学 | A kind of simple and efficient 3 D human body method for reconstructing based on single Kinect |
| CN106203335B (en) * | 2016-07-11 | 2019-05-28 | 厦门大学 | Sign board visibility evaluation method based on three-dimensional point cloud |
| CN106600639B (en) * | 2016-12-09 | 2019-06-25 | 江南大学 | The ICP pose location technology of genetic algorithm combining adaptive threshold value constraint |
| CN106846386B (en) * | 2017-02-08 | 2019-06-28 | 南通大学 | 3D cranium method for registering images based on ROI and conformal geometric algebra property invariant |
| FI127555B (en) | 2017-04-05 | 2018-08-31 | Oy Mapvision Ltd | Machine vision system with coordinate correction |
| CN107203269B (en) * | 2017-06-01 | 2020-05-05 | 珠海智康医疗科技有限公司 | Data processing method and device |
| CN107665496B (en) * | 2017-08-25 | 2020-04-10 | 北京控制工程研究所 | Three-dimensional attitude registration method |
| CN107590829B (en) * | 2017-09-18 | 2020-06-30 | 西安电子科技大学 | Seed point picking method suitable for multi-view dense point cloud data registration |
| CN107633539A (en) * | 2017-09-25 | 2018-01-26 | 潍坊学院 | A kind of three-dimensional point cloud model data compression method based on four side patch divisions |
| CN107886529B (en) * | 2017-12-06 | 2020-04-10 | 重庆理工大学 | Point cloud registration method for three-dimensional reconstruction |
| CN108257163B (en) * | 2017-12-08 | 2020-04-07 | 西安电子科技大学 | Two-point cloud registration method under known scanning point position |
| CN109965979A (en) * | 2017-12-27 | 2019-07-05 | 上海复旦数字医疗科技股份有限公司 | A kind of steady Use of Neuronavigation automatic registration method without index point |
| CN108230402B (en) * | 2018-01-23 | 2021-09-21 | 北京易智能科技有限公司 | Three-dimensional calibration method based on triangular pyramid model |
| CN108253911B (en) * | 2018-01-29 | 2019-10-11 | 西南交通大学 | A kind of workpiece pose method of adjustment based on measurement point geometrical characteristic iteration registration |
| CN108731615B (en) * | 2018-03-19 | 2020-12-25 | 苏州玻色智能科技有限公司 | Equipment and method for detecting curved glass panel |
| CN108765474A (en) * | 2018-04-17 | 2018-11-06 | 天津工业大学 | A kind of efficient method for registering for CT and optical scanner tooth model |
| CN109523578A (en) * | 2018-08-27 | 2019-03-26 | 中铁上海工程局集团有限公司 | A kind of matching process of bim model and point cloud data |
| CN110160502B (en) | 2018-10-12 | 2022-04-01 | 腾讯科技（深圳）有限公司 | Map element extraction method, device and server |
| CN111368852A (en) * | 2018-12-26 | 2020-07-03 | 沈阳新松机器人自动化股份有限公司 | Article identification and pre-sorting system and method based on deep learning and robot |
| CN109489553B (en) * | 2018-12-27 | 2020-10-16 | 中国科学院长春光学精密机械与物理研究所 | Method, device, equipment and storage medium for generating space marker point library |
| CN109781010B (en) * | 2019-01-21 | 2020-11-10 | 珠海博明软件有限公司 | Point cloud data calculation method, device and system for large-range data splicing |
| CN109765845A (en) * | 2019-03-07 | 2019-05-17 | 台州稳迅自动化设备有限公司 | A kind of numerical control cutting machine workpiece calibrates precisely controlling system automatically |
| CN110415361B (en) * | 2019-07-26 | 2020-05-15 | 北京罗森博特科技有限公司 | Method and device for splicing broken objects |
| CN110559075B (en) * | 2019-08-05 | 2021-09-24 | 常州锦瑟医疗信息科技有限公司 | Intraoperative augmented reality registration method and device |
| CN110473239A (en) * | 2019-08-08 | 2019-11-19 | 刘秀萍 | A kind of high-precision point cloud registration method of 3 D laser scanning |
| CN111369603B (en) * | 2020-02-25 | 2023-09-29 | 阿波罗智能技术(北京)有限公司 | Evaluation method and device of point cloud registration technology, electronic equipment and readable storage medium |
| CN111445540B (en) * | 2020-03-26 | 2023-04-18 | 重庆理工大学 | Automatic registration method for RGB colored three-dimensional point cloud |
| CN111816282A (en) * | 2020-07-01 | 2020-10-23 | 杭州脉流科技有限公司 | Left ventricle functional analysis method and device based on CT image, computer equipment and storage medium |
| CN111882614A (en) * | 2020-07-30 | 2020-11-03 | 南京溧航仿生产业研究院有限公司 | KNN-ICP algorithm-based free-form surface positioning method |
| CN112015138B (en) * | 2020-09-04 | 2021-08-10 | 中国工程物理研究院机械制造工艺研究所 | Blade contour error evaluation method based on K nearest neighbor iterative nearest grid algorithm |
| CN113409242A (en) * | 2021-02-23 | 2021-09-17 | 杭州哲域智能科技有限公司 | Intelligent monitoring method for point cloud of rail intersection bow net |
| CN113284252B (en) * | 2021-03-03 | 2023-05-05 | 山东交通学院 | Resetting method and resetting system of fractured bone model based on VTK |
| CN113362446B (en) * | 2021-05-25 | 2023-04-07 | 上海奥视达智能科技有限公司 | Method and device for reconstructing object based on point cloud data |
| CN113284170A (en) * | 2021-05-26 | 2021-08-20 | 北京智机科技有限公司 | Point cloud rapid registration method |
| CN114111638B (en) * | 2021-09-07 | 2024-02-20 | 领先光学技术(江苏)有限公司 | Curved surface detection method based on phase deflection |
| CN116664575B (en) * | 2023-07-31 | 2023-11-03 | 厦门微图软件科技有限公司 | Defect detection method, device and equipment based on point cloud registration |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7602963B2 (en) * | 2006-01-10 | 2009-10-13 | General Electric Company | Method and apparatus for finding anomalies in finished parts and/or assemblies |
| CN103150747A (en) * | 2012-11-16 | 2013-06-12 | 佳都新太科技股份有限公司 | Point cloud registration algorithm based on topological characteristic |
| CN103236064B (en) * | 2013-05-06 | 2016-01-13 | 东南大学 | A kind of some cloud autoegistration method based on normal vector |
-
2014
- 2014-05-16 CN CN201410208338.6A patent/CN103955939B/en not_active Expired - Fee Related
Non-Patent Citations (2)
| Title |
|---|
| "ICP算法在点云配准中的应用";戴静兰等;《中国图象图形学报》;20070531;第12卷(第3期);第517页摘要及第1节第1段，第518-519页第3节第1段及第3.1节第1-4段，第519-520页第3.2节第3-4段 * |
| "Study on Edge Detection of LIDAR Point Cloud";Rongrong Wang etal;《Proceedings of 2011 International Conference on Intelligent Computation and Bio-Medical Instrumentation》;20111231;第II节第A小节第1段，第III节第A小节第1-2段 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103955939A (en) | 2014-07-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103955939B (en) | 3 D scanning system midpoint cloud edge feature point method for registering | |
| CN109767463A (en) | A kind of three-dimensional point cloud autoegistration method | |
| CN104897062B (en) | Visual measurement method and device for shape and position deviation of part non-coplanar parallel holes | |
| CN109544612B (en) | Point cloud registration method based on feature point geometric surface description | |
| CN108830902A (en) | A kind of workpiece identification at random and localization method based on points cloud processing | |
| Kanhangad et al. | Contactless and pose invariant biometric identification using hand surface | |
| EP3091479B1 (en) | Method and apparatus for fingerprint identification | |
| CN104484508B (en) | Optimizing method for noncontact three-dimensional matching detection of complex curved-surface part | |
| CN105046694B (en) | A kind of point cloud rapid registering method based on surface fitting coefficient characteristics | |
| CN105783786B (en) | Part chamfering measuring method and device based on structured light vision | |
| CN110473239A (en) | A kind of high-precision point cloud registration method of 3 D laser scanning | |
| CN113706537B (en) | Irregular workpiece detection method based on machine vision | |
| CN106327536B (en) | A kind of collar measurement method based on section point cloud | |
| CN112257722B (en) | Point cloud fitting method based on robust nonlinear Gaussian-Hermer model | |
| CN113516695B (en) | Point cloud registration strategy in laser profiler flatness measurement | |
| CN107133966A (en) | A kind of three-dimensional sonar image background dividing method based on sampling consistency algorithm | |
| CN104134058B (en) | A kind of face image processing process | |
| CN109472816A (en) | A kind of point cloud registration method | |
| CN108389260A (en) | A kind of three-dimensional rebuilding method based on Kinect sensor | |
| Arachchige et al. | Intensity integrated Laplacian-based thickness measurement for detecting human metaphase chromosome centromere location | |
| CN104764407A (en) | Method for measuring thickness of cable protecting bush accurately | |
| CN107633507A (en) | LCD defect inspection methods based on contour detecting and characteristic matching | |
| CN114648445B (en) | Multi-view high-resolution point cloud splicing method based on feature point extraction and fine registration optimization | |
| CN113936303B (en) | Method for determining maximum inscribed rectangle of hand image and image identification method | |
| Andreadis et al. | Generalized digital reassembly using geometric registration |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20180619 Termination date: 20190516 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |