CN109872354A - Multi-angle of view point cloud registration method and system based on nonlinear optimization - Google Patents
Multi-angle of view point cloud registration method and system based on nonlinear optimization Download PDFInfo
- Publication number
- CN109872354A CN109872354A CN201910079082.6A CN201910079082A CN109872354A CN 109872354 A CN109872354 A CN 109872354A CN 201910079082 A CN201910079082 A CN 201910079082A CN 109872354 A CN109872354 A CN 109872354A
- Authority
- CN
- China
- Prior art keywords
- rigid body
- body translation
- cloud
- translation parameter
- point cloud
- 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.)
- Pending
Links
Landscapes
- Image Processing (AREA)
Abstract
The present invention provides a kind of multi-angle of view point cloud registration method based on nonlinear optimization, comprising: obtains the point cloud data at multiple and different visual angles to be matched, and obtains for each point cloud data to be transformed into the corresponding initial rigid body translation parameter of world coordinate system；When initial rigid body translation parameter meets preset condition, corresponding point cloud data is transformed under world coordinate system according to initial rigid body translation parameter and carries out just registration；Determine the corresponding points after being just registrated between each cloud and other clouds；Objective function is constructed according to the Euclidean distance minimum between all corresponding points and carries out global optimization, to solve the optimal rigid body translation parameter of each cloud；Accuracy registration is carried out to point cloud data by optimal rigid body translation parameter.The present invention also provides a kind of multi-angle of view point cloud registering systems.The present invention carries out the shortcomings that global optimization solves optimal rigid body translation parameter, the prior art is overcome to be easily trapped into locally optimal solution by construction objective function, and registration accuracy is higher.
Description
[technical field]
The invention belongs to method for three-dimensional measurement and Instrument technology field more particularly to a kind of more views based on nonlinear optimization
Angle point cloud registration method and system.
[background technique]
When carrying out 3-D scanning to object, it usually needs be scanned from the different direction of measured object more complete to obtain
Three-dimensional data, and need to the multi-view angle three-dimensional point cloud of gained measured object carry out registration be in the same coordinate system.It is existing
Multi-angle of view point cloud matching technology be using ICP (Iterative Closest Point, iteration closest approach) algorithm carry out
To match, ICP is carried out by the point cloud that two different perspectivess are chosen in combination, this process of repetition is multiple, until whole convergences, thus
To the matching result of multi-angle of view point cloud entirety.
And when using ICP matching, do not ensure that multi-angle of view point cloud ICP matching restrains that obtain global registration optimal simultaneously
As a result, self judgment convergence direction and convergence step-length are unable to during iteration convergence, when the initial position of to be matched cloud
In the case that corresponding points number is less between unreasonable or point cloud, matching optimization result is easily trapped into local optimum.In consideration of it, real
The multi-angle of view point cloud registration method and system that it is necessary to provide a kind of based on nonlinear optimization are to overcome drawbacks described above.
[summary of the invention]
The present invention propose it is a kind of can global optimization to improve the multi-angle of view point cloud based on nonlinear optimization of registration accuracy
Method for registering and system.
To achieve the goals above, the present invention provides a kind of multi-angle of view point cloud registration method based on nonlinear optimization, packet
Include following steps:
The point cloud data at multiple and different visual angles to be matched is obtained, and is obtained for each point cloud data to be transformed into generation
The corresponding initial rigid body translation parameter of boundary's coordinate system；
It, will be corresponding according to the initial rigid body translation parameter when the initial rigid body translation parameter meets preset condition
Point cloud data, which is transformed under world coordinate system, carries out just registration；
Determine the corresponding points after being just registrated between each cloud and other clouds；
Objective function is constructed according to the Euclidean distance minimum between all corresponding points carries out global optimization, it is each to solve
The optimal rigid body translation parameter of point cloud；
Accuracy registration is carried out to the point cloud data by the optimal rigid body translation parameter.
In a preferred embodiment, the determination is corresponding between each cloud and other clouds after being just registrated
Point the step of include:
Obtain the point and certain point nearest with point distance in other clouds in cloud；
When a point of cloud in other clouds certain point at a distance from recently and be less than distance threshold when, then be judged as
Corresponding points.
In a preferred embodiment, the Euclidean distance minimum according between all corresponding points constructs target letter
The step of number carries out global optimizations, optimal rigid body translation parameter to solve each cloud, comprising:
So that objective function to be optimized is reached minimum by global iterative, intermediate rigid body translation parameter is calculated；
Posture tune is carried out to the point cloud data being transformed under world coordinate system according to the intermediate rigid body translation parameter
It is whole；
Judge Euclidean distance after adjustment between all corresponding points whether global convergence；
When Euclidean distance after judgement is adjusted between all corresponding points meets global convergence, it is determined that the centre
Rigid body translation parameter is optimal rigid body translation parameter.
In a preferred embodiment, the method also includes: after judgement is adjusted between all corresponding points
When Euclidean distance does not meet global convergence, using the intermediate rigid body translation parameter initial rigid body new as global iterative next time
Transformation parameter simultaneously enters global iterative next time, to solve optimal rigid body translation parameter.
In a preferred embodiment, the objective function isOijExpression formula beWherein, norm2 is two norms, OijIndicate pair of i-th cloud and j-th cloud
Euclidean distance between should putting, RTi、RTjThe rigid body translation parameter of respectively i-th cloud and j-th cloud, pik、pjkRespectively
Indicate i-th cloud with the kth on j-th cloud to corresponding point data, nikIndicate k-th of corresponding points on i-th cloud
Normal vector.
In a preferred embodiment, the preset condition is that iteration residual error is less than a reference value.
To achieve the goals above, the present invention provides a kind of multi-angle of view point cloud registering system based on nonlinear optimization, packet
It includes:
Module is obtained, for obtaining the point cloud data at multiple and different visual angles to be matched, and is obtained for by each point
Cloud data are transformed into the corresponding initial rigid body translation parameter of world coordinate system；
First registration module, for when the initial rigid body translation parameter meets preset condition, according to initial rigid body translation
Corresponding point cloud data is transformed under world coordinate system and carries out just registration by parameter；
Corresponding points determining module, for determining the corresponding points after being just registrated between each cloud and other clouds；
Parametric solution module carries out entirely for constructing objective function according to the Euclidean distance minimum between all corresponding points
Office's optimization, to solve the optimal rigid body translation parameter of each cloud；
Smart registration module, for carrying out accuracy registration to the point cloud data by the optimal rigid body translation parameter.
In a preferred embodiment, the corresponding points determining module includes:
Corresponding points obtain module, for obtains a point in one cloud and in other clouds it is nearest with point distance
Certain point；
Corresponding points judgment module, for when cloud a point in other clouds at a distance from certain point it is nearest and be less than
When distance threshold, it is judged as corresponding points.
In a preferred embodiment, the parametric solution module includes:
Intermediate rigid body is calculated for making objective function to be optimized reach minimum by global iterative in computing module
Transformation parameter；
Module is adjusted, for carrying out appearance to the point cloud data being transformed under world coordinate system according to intermediate rigid body translation parameter
State adjustment；
Condition judgment module, for judge Euclidean distance after adjustment between all corresponding points whether global convergence；
First judging submodule, for the Europe between corresponding points all after condition judgment module judgement is adjusted
When family name's distance meets global convergence, it is determined that the intermediate rigid body translation parameter is optimal rigid body translation parameter.
In a preferred embodiment, the parametric solution module further includes second judgment submodule；Described second sentences
Disconnected submodule is not met entirely for the Euclidean distance between corresponding points all after condition judgment module judgement is adjusted
When office's convergence, using the intermediate rigid body translation parameter initial rigid body translation parameter new as global iterative next time and under entering
Global iterative, to solve optimal rigid body translation parameter.
Compared with prior art, the beneficial effects of the present invention are: obtained by that will put under Cloud transform to world coordinate system
Initial rigid body translation parameter constructs objective function according to the Euclidean distance minimum between all corresponding points and carries out global optimization,
It is further solved in conjunction with initial rigid body translation parameter and point cloud data, obtains optimal rigid body translation parameter, overcome the prior art
The shortcomings that being easily trapped into locally optimal solution, so that the registration result precision of multi-angle of view point cloud is higher.
[Detailed description of the invention]
In order to more clearly explain the embodiment of the invention or the technical proposal in the existing technology, to embodiment or will show below
There is attached drawing needed in technical description to be briefly described, it should be apparent that, the accompanying drawings in the following description is only this
Some embodiments of invention for those of ordinary skill in the art without creative efforts, can be with
It obtains other drawings based on these drawings.
Fig. 1 is the process for the multi-angle of view point cloud registration method based on nonlinear optimization that present pre-ferred embodiments provide
Figure；
Fig. 2 is the flow diagram of step S103 shown in Fig. 1；
Fig. 3 is the flow diagram of step S104 shown in Fig. 1；
Fig. 4 is the structural frames for the multi-angle of view point cloud registering system based on nonlinear optimization that present pre-ferred embodiments provide
Figure.
[specific embodiment]
It is clear in order to be more clear the purpose of the present invention, technical solution and advantageous effects, below in conjunction with attached drawing and
Specific embodiment, the present invention will be described in further detail.It should be understood that specific implementation described in this specification
Mode is not intended to limit the present invention just for the sake of explaining the present invention.
Referring to Fig. 1, the present invention provides a kind of multi-angle of view point cloud registration method based on nonlinear optimization, including following step
It is rapid:
Step S101, obtains the point cloud data at multiple and different visual angles to be matched, and obtains for by each cloud number
According to being transformed into the corresponding initial rigid body translation parameter of world coordinate system.
In step s101, different scanning modes is selected according to scene to obtain the multiple and different visual angles of object to be scanned
Point cloud, specifically surrounds object motion scan instrument to be scanned centered on object to be scanned, surrounds object to be scanned from multiple directions
Scanning, each direction obtain one section of point cloud, guarantee that there are overlapping regions for adjacent two sections of points cloud.Generally for obtaining testee
Complete data model needs to be registrated point cloud data, and process is that the point set obtained from each visual angle is merged into unification
Coordinate system under form a complete point cloud, then can facilitate carry out visualized operation.Point cloud data includes three-dimensional sits
Mark, laser reflection intensity and colouring information etc., the point cloud data obtained in the present invention is three-dimensional coordinate.It, will be each in this step
The point cloud at visual angle, which is transformed under world coordinate system, carries out initial transformation, to obtain initial rigid body translation parameter.
Step S102 will be corresponded to when initial rigid body translation parameter meets preset condition according to initial rigid body translation parameter
Point cloud data be transformed under world coordinate system carry out just registration.
In step s 102, initial rigid body translation parameter is usually in pose of camera (motion estimation) and to minimize mistake
The rigid body translation estimated under module, needs to check whether initial rigid body translation parameter meets preset condition, to make
All the points cloud under world coordinate system is substantially matching.In the present embodiment, preset condition is that iteration residual error is less than a reference value, when repeatedly
When excessive for residual error, then judge that initial rigid body translation parameter is unreasonable, needs to repeat step S101.
Step S103 determines the corresponding points after being just registrated between each cloud and other clouds.
Referring to Fig. 2, step S103 further comprises following sub-step:
Step S201 obtains a point and certain point nearest with point distance in other clouds in cloud.
In step s 201, by closely located with other clouds in the available each cloud of the point cloud data of acquisition
Point pair.In present embodiment, point to the distance between can be Euclidean distance or the relevant other distances of attribute.
Step S202, when in cloud a point in other clouds certain point at a distance from recently and be less than distance threshold
When, then it is judged as corresponding points.
In step S202, above-mentioned distance threshold can be set as needed, to remove the corresponding points of erroneous matching.This reality
It applies in mode, distance threshold may be set to 10 millimeters, i.e., as a point p in cloudiWith certain point p in other cloudsjAway from
From recently and when less than 10 millimeters, piWith pjFor corresponding points.
Step S104 constructs objective function according to the Euclidean distance minimum between all corresponding points and carries out global optimization,
To solve the optimal rigid body translation parameter of each cloud.
In step S104, the purpose of global optimization is adjusted to initial rigid body translation parameter, obtains optimal rigid body
Transformation parameter, can avoid occur in the prior art by ICP match fall into local optimum as a result, simultaneously in point cloud to be matched
Initial position it is unreasonable or point cloud between corresponding points number it is less in the case where, registration result is also not readily susceptible to influence.
Referring to Fig. 3, step S104 further comprises following sub-step:
Step S301 makes objective function to be optimized reach minimum, intermediate rigid body translation is calculated by global iterative
Parameter.
In the present embodiment, the expression formula of above-mentioned objective function isThe substitute point in objective function
Cloud data and initial rigid body translation parameter, continuous iteration is minimum until the Euclidean distance of the corresponding points between all the points cloud,
Intermediate rigid body translation parameter can be solved after optimizing.
Specifically, OijExpression formula it is as follows:
Wherein, norm2 is two norms, OijIndicate the Euclidean distance between i-th cloud and the corresponding points of j-th cloud,
RTi、RTjThe rigid body translation parameter of respectively i-th cloud and j-th cloud, pik、pjkRespectively indicate i-th cloud and j-th
Kth on point cloud is to corresponding point data, nikIndicate the normal vector of k-th of corresponding points on i-th cloud.
Step S302 carries out posture tune to the point cloud data being transformed under world coordinate system according to intermediate rigid body translation parameter
It is whole.
In step s 302, according to intermediate rigid body translation parameter (including spin matrix and translation vector) to point cloud data into
Row adjustment carries out global optimization to all the points cloud after first registration.
Step S303, judge Euclidean distance after adjustment between all corresponding points whether global convergence.
In step S303, to judging whether the Euclidean distance between corresponding points restrains again after all the points cloud global optimization,
If meeting global convergence, S304 is entered step, terminates iterative optimization procedure；Otherwise, S305 is entered step, i.e. repeatedly step
S201 re-starts global optimization.
Step S304, when the Euclidean distance after judgement is adjusted between all corresponding points meets global convergence, then really
Fixed intermediate rigid body translation parameter is optimal rigid body translation parameter.
Step S305 will when the Euclidean distance after judgement is adjusted between all corresponding points does not meet global convergence
The intermediate rigid body translation parameter initial rigid body translation parameter new as global iterative next time simultaneously enters global iterative next time, with
Solve optimal rigid body translation parameter.
In step S305, when the Euclidean distance between all corresponding points is unsatisfactory for global convergence condition, then into next
Secondary global iterative meets the iteration of global convergence by multiple loop iteration until reaching the Euclidean distance between all corresponding points
Number, to solve optimal rigid body translation parameter.It should be noted that being obtained during the S times global iterative (wherein S >=2)
The initial rigid body translation parameter taken is the S-1 times intermediate rigid body translation parameter being calculated, and according to the S-1 times pose adjustment
Point cloud afterwards redefines the corresponding points of cloud and other clouds at each, according further to the Euclidean between all corresponding points away from
Objective function is constructed from minimum and carries out global optimization, to solve the optimal rigid body translation parameter of each cloud.
Step S105 carries out accuracy registration to the point cloud data by optimal rigid body translation parameter.
In step s105, optimal rigid body translation parameter meets the condition of convergence, than initial rigid body translation parameter error more
It is small, the matching target that the registration of cloud is optimal is put, obtained Whole Data Model precision is higher.
Multi-angle of view point cloud registration method provided in an embodiment of the present invention based on nonlinear optimization, by point Cloud transform to the world
Initial rigid body translation parameter is obtained under coordinate system, constructed according to the Euclidean distance minimum between all corresponding points objective function into
Row global optimization is further solved in conjunction with initial rigid body translation parameter and point cloud data, is obtained optimal rigid body translation parameter, is overcome
The shortcomings that prior art is easily trapped into locally optimal solution, so that the registration result precision of multi-angle of view point cloud is higher.
On the other hand, referring to Fig. 4, being based on above method embodiment, the present invention also provides one kind to be based on nonlinear optimization
Multi-angle of view point cloud registering system 100, specifically include:
Module 10 is obtained, for obtaining the point cloud data at multiple and different visual angles to be matched, and is obtained for will be each
Point cloud data is transformed into the corresponding initial rigid body translation parameter of world coordinate system.
First registration module 20, for being joined according to initial rigid body translation when initial rigid body translation parameter meets preset condition
Corresponding point cloud data is transformed under world coordinate system and carries out just registration by number.
Corresponding points determining module 30, for determining the corresponding points after being just registrated between each cloud and other clouds.
Parametric solution module 40 is carried out for constructing objective function according to the Euclidean distance minimum between all corresponding points
Global optimization, to solve the optimal rigid body translation parameter of each cloud.
Smart registration module 50, for carrying out accuracy registration to the point cloud data by optimal rigid body translation parameter.
In present embodiment, above-mentioned corresponding points determining module 30 further comprises:
Corresponding points obtain module 31, for obtains a point in one cloud and in other clouds with the point apart from most
Certain close point.
Corresponding points judgment module 32, for when each cloud a point in other clouds certain put at a distance from it is nearest and small
When distance threshold, it is judged as corresponding points.
In present embodiment, above-mentioned parametric solution module 40 further comprises:
Computing module 41 is calculated intermediate rigid for making objective function to be optimized reach minimum by global iterative
Body transformation parameter.
Module 42 is adjusted, for carrying out according to intermediate rigid body translation parameter to the point cloud data being transformed under world coordinate system
Pose adjustment.
Condition judgment module 43, for judge Euclidean distance after adjustment between all corresponding points whether global convergence.
First judging submodule 44, for the Europe between corresponding points all after condition judgment module 43 judges to be adjusted
When family name's distance meets global convergence, it is determined that intermediate rigid body translation parameter is optimal rigid body translation parameter.
Specifically, above-mentioned parametric solution module 40 further includes second judgment submodule 45.Second judgment submodule 45 is used
When the Euclidean distance after condition judgment module 43 judges to be adjusted between all corresponding points does not meet global convergence, will in
Between the rigid body translation parameter initial rigid body translation parameter new as global iterative next time and enter global iterative next time, in the hope of
Solve optimal rigid body translation parameter.
Those of ordinary skill in the art will appreciate that all or part of the steps in the various methods of above-described embodiment is can
It is completed with instructing relevant hardware by program, which can be stored in a computer readable storage medium, storage
Medium may include: read-only memory (ROM, Read Only Memory), random access memory (RAM, Random
Access Memory), disk or CD etc..
The foregoing is merely illustrative of the preferred embodiments of the present invention, is to combine specific preferred embodiment to institute of the present invention
The further description of work is, and it cannot be said that specific implementation of the invention is confined to these explanations.It is all in spirit of the invention and
Made any modifications, equivalent replacements, and improvements etc., should be included within the scope of the present invention within principle.
Claims (10)
1. a kind of multi-angle of view point cloud registration method based on nonlinear optimization, which comprises the following steps:
The point cloud data at multiple and different visual angles to be matched is obtained, and obtains and is sat for each point cloud data to be transformed into the world
The corresponding initial rigid body translation parameter of mark system；
When the initial rigid body translation parameter meets preset condition, according to the initial rigid body translation parameter by corresponding cloud
Data, which are transformed under world coordinate system, carries out just registration；
Determine the corresponding points after being just registrated between each cloud and other clouds；
Objective function is constructed according to the Euclidean distance minimum between all corresponding points and carries out global optimization, to solve each cloud
Optimal rigid body translation parameter；
Accuracy registration is carried out to the point cloud data by the optimal rigid body translation parameter.
2. the multi-angle of view point cloud registration method based on nonlinear optimization as described in claim 1, which is characterized in that the determination
The step of corresponding points after being just registrated between each cloud and other clouds includes:
Obtain the point and certain point nearest with point distance in other clouds in cloud；
When a point of cloud in other clouds certain point at a distance from recently and be less than distance threshold when, then be judged as corresponding
Point.
3. the multi-angle of view point cloud registration method based on nonlinear optimization as described in claim 1, which is characterized in that the basis
Euclidean distance minimum between all corresponding points carries out global optimization to construct objective function, to solve the optimal rigid of each cloud
The step of body transformation parameter, comprising:
So that objective function to be optimized is reached minimum by global iterative, intermediate rigid body translation parameter is calculated；
Pose adjustment is carried out to the point cloud data being transformed under world coordinate system according to the intermediate rigid body translation parameter；
Judge Euclidean distance after adjustment between all corresponding points whether global convergence；
When Euclidean distance after judgement is adjusted between all corresponding points meets global convergence, it is determined that the intermediate rigid body
Transformation parameter is optimal rigid body translation parameter.
4. the multi-angle of view point cloud registration method based on nonlinear optimization as claimed in claim 3, which is characterized in that the method
Further include: it is when the Euclidean distance after judgement is adjusted between all corresponding points does not meet global convergence, the centre is rigid
The body transformation parameter initial rigid body translation parameter new as global iterative next time simultaneously enters global iterative next time, to solve most
Excellent rigid body translation parameter.
5. the multi-angle of view point cloud registration method based on nonlinear optimization as claimed in claim 3, which is characterized in that the target
Function isOijExpression formula beWherein, norm2 is two norms,
OijIndicate the Euclidean distance between i-th cloud and the corresponding points of j-th cloud, RTi、RTjRespectively i-th cloud and jth
The rigid body translation parameter of a cloud, pik、pjkI-th cloud is respectively indicated with the kth on j-th cloud to corresponding point data, nik
Indicate the normal vector of k-th of corresponding points on i-th cloud.
6. the multi-angle of view point cloud registration method based on nonlinear optimization as described in claim 1, which is characterized in that described default
Condition is that iteration residual error is less than a reference value.
7. a kind of multi-angle of view point cloud registering system based on nonlinear optimization characterized by comprising
Module is obtained, for obtaining the point cloud data at multiple and different visual angles to be matched, and is obtained for by each cloud number
According to being transformed into the corresponding initial rigid body translation parameter of world coordinate system；
First registration module, for when the initial rigid body translation parameter meets preset condition, according to initial rigid body translation parameter
Corresponding point cloud data is transformed under world coordinate system and carries out just registration；
Corresponding points determining module, for determining the corresponding points after being just registrated between each cloud and other clouds；
Parametric solution module, it is global excellent for constructing objective function progress according to the Euclidean distance minimum between all corresponding points
Change, to solve the optimal rigid body translation parameter of each cloud；
Smart registration module, for carrying out accuracy registration to the point cloud data by the optimal rigid body translation parameter.
8. the multi-angle of view point cloud registering system based on nonlinear optimization as claimed in claim 7, which is characterized in that the correspondence
Putting determining module includes:
Corresponding points obtain module, for obtains a point in one cloud and in other clouds with point distance recently certain
Point；
Corresponding points judgment module, for when cloud a point in other clouds at a distance from certain point recently and less than distance
When threshold value, it is judged as corresponding points.
9. the multi-angle of view point cloud registering system based on nonlinear optimization as claimed in claim 7, which is characterized in that the parameter
Solving module includes:
Intermediate rigid body translation is calculated for making objective function to be optimized reach minimum by global iterative in computing module
Parameter；
Module is adjusted, for carrying out posture tune to the point cloud data being transformed under world coordinate system according to intermediate rigid body translation parameter
It is whole；
Condition judgment module, for judge Euclidean distance after adjustment between all corresponding points whether global convergence；
First judging submodule, for Euclidean after condition judgment module judgement is adjusted between all corresponding points away from
When from meeting global convergence, it is determined that the intermediate rigid body translation parameter is optimal rigid body translation parameter.
10. the multi-angle of view point cloud registering system based on nonlinear optimization as claimed in claim 9, which is characterized in that the ginseng
It further includes second judgment submodule that number, which solves module,；The second judgment submodule, for judging when the condition judgment module
When Euclidean distance after being adjusted between all corresponding points does not meet global convergence, using the intermediate rigid body translation parameter as
The new initial rigid body translation parameter of global iterative and enter global iterative next time next time, to solve optimal rigid body translation ginseng
Number.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910079082.6A CN109872354A (en) | 2019-01-28 | 2019-01-28 | Multi-angle of view point cloud registration method and system based on nonlinear optimization |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910079082.6A CN109872354A (en) | 2019-01-28 | 2019-01-28 | Multi-angle of view point cloud registration method and system based on nonlinear optimization |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109872354A true CN109872354A (en) | 2019-06-11 |
Family
ID=66918213
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910079082.6A Pending CN109872354A (en) | 2019-01-28 | 2019-01-28 | Multi-angle of view point cloud registration method and system based on nonlinear optimization |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109872354A (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111612887A (en) * | 2020-04-30 | 2020-09-01 | 北京的卢深视科技有限公司 | Human body measuring method and device |
| CN113192114A (en) * | 2021-07-01 | 2021-07-30 | 四川大学 | Blade multi-field point cloud registration method based on overlapping features and local distance constraint |
| CN114004878A (en) * | 2020-07-28 | 2022-02-01 | 株式会社理光 | Alignment device, alignment method, alignment system, storage medium, and computer device |
| CN114820955A (en) * | 2022-06-30 | 2022-07-29 | 苏州魔视智能科技有限公司 | Symmetric plane completion method, device, equipment and storage medium |
| WO2022165876A1 (en) * | 2021-02-06 | 2022-08-11 | 湖南大学 | Wgan-based unsupervised multi-view three-dimensional point cloud joint registration method |
| CN116299367A (en) * | 2023-05-18 | 2023-06-23 | 中国测绘科学研究院 | Multi-laser space calibration method |
-
2019
- 2019-01-28 CN CN201910079082.6A patent/CN109872354A/en active Pending
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111612887A (en) * | 2020-04-30 | 2020-09-01 | 北京的卢深视科技有限公司 | Human body measuring method and device |
| CN114004878A (en) * | 2020-07-28 | 2022-02-01 | 株式会社理光 | Alignment device, alignment method, alignment system, storage medium, and computer device |
| WO2022165876A1 (en) * | 2021-02-06 | 2022-08-11 | 湖南大学 | Wgan-based unsupervised multi-view three-dimensional point cloud joint registration method |
| CN112837356B (en) * | 2021-02-06 | 2024-05-31 | 湖南大学 | WGAN-based non-supervision multi-view three-dimensional point cloud joint registration method |
| CN113192114A (en) * | 2021-07-01 | 2021-07-30 | 四川大学 | Blade multi-field point cloud registration method based on overlapping features and local distance constraint |
| CN113192114B (en) * | 2021-07-01 | 2021-09-03 | 四川大学 | Blade multi-field point cloud registration method based on overlapping features and local distance constraint |
| CN114820955A (en) * | 2022-06-30 | 2022-07-29 | 苏州魔视智能科技有限公司 | Symmetric plane completion method, device, equipment and storage medium |
| CN114820955B (en) * | 2022-06-30 | 2022-11-18 | 苏州魔视智能科技有限公司 | Symmetric plane completion method, device, equipment and storage medium |
| CN116299367A (en) * | 2023-05-18 | 2023-06-23 | 中国测绘科学研究院 | Multi-laser space calibration method |
| CN116299367B (en) * | 2023-05-18 | 2024-01-26 | 中国测绘科学研究院 | Multi-laser space calibration method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109872354A (en) | Multi-angle of view point cloud registration method and system based on nonlinear optimization | |
| CN109682382B (en) | Global fusion positioning method based on self-adaptive Monte Carlo and feature matching | |
| CN104063894B (en) | Point cloud three-dimensional model reestablishing method and system | |
| CN105096386B (en) | A wide range of complicated urban environment geometry map automatic generation method | |
| CN110363858A (en) | A kind of three-dimensional facial reconstruction method and system | |
| Grisetti et al. | Robust optimization of factor graphs by using condensed measurements | |
| CN106023298B (en) | Point cloud Rigid Registration method based on local Poisson curve reestablishing | |
| CN106153048A (en) | A kind of robot chamber inner position based on multisensor and Mapping System | |
| CN103868510B (en) | A kind of quickly whole day star chart autonomous fixed star recognition methods | |
| CN110147815A (en) | Multiframe point cloud fusion method and device based on K mean cluster | |
| CN107122705A (en) | Face critical point detection method based on three-dimensional face model | |
| CN108090443A (en) | Scene text detection method and system based on deeply study | |
| CN110524544A (en) | A kind of control method of manipulator motion, terminal and readable storage medium storing program for executing | |
| CN105701455A (en) | Active shape model (ASM) algorithm-based face characteristic point acquisition and three dimensional face modeling method | |
| CN109598781A (en) | The method of puppet 3D frame is obtained from 2D bounding box by regression analysis and uses the learning device and test device of this method | |
| KR102608473B1 (en) | Method and apparatus for aligning 3d model | |
| CN111427047A (en) | Autonomous mobile robot S L AM method in large scene | |
| CN109460267A (en) | Mobile robot offline map saves and real-time method for relocating | |
| Einhorn et al. | Generic 2D/3D SLAM with NDT maps for lifelong application | |
| CN109872366A (en) | Object dimensional method for detecting position and device based on depth fitting degree assessment network | |
| CN106846416A (en) | Unit beam splitting bi-eye passiveness stereo vision Accurate Reconstruction and subdivision approximating method | |
| CN113108773A (en) | Grid map construction method integrating laser and visual sensor | |
| CN103093237B (en) | A kind of method for detecting human face of structure based model | |
| CN109242769A (en) | A kind of image processing method and device | |
| CN109993108B (en) | Gesture error correction method, system and device under a kind of augmented reality environment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |