CN107784656B - Part point cloud segmentation method based on geometric elements - Google Patents
Part point cloud segmentation method based on geometric elements Download PDFInfo
- Publication number
- CN107784656B CN107784656B CN201710852830.0A CN201710852830A CN107784656B CN 107784656 B CN107784656 B CN 107784656B CN 201710852830 A CN201710852830 A CN 201710852830A CN 107784656 B CN107784656 B CN 107784656B
- Authority
- CN
- China
- Prior art keywords
- point
- point cloud
- pile
- elements
- points
- 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.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T15/00—3D [Three Dimensional] image rendering
- G06T15/10—Geometric effects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/30—Determination of transform parameters for the alignment of images, i.e. image registration
- G06T7/33—Determination of transform parameters for the alignment of images, i.e. image registration using feature-based methods
- G06T7/344—Determination of transform parameters for the alignment of images, i.e. image registration using feature-based methods involving models
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10028—Range image; Depth image; 3D point clouds
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Geometry (AREA)
- Computer Graphics (AREA)
- Length Measuring Devices With Unspecified Measuring Means (AREA)
- Image Generation (AREA)
Abstract
The invention discloses a part point cloud segmentation method based on geometric elements, which mainly comprises the following steps: establishing a part model in an IGES format and establishing a part point cloud. Registering the part model and the part point cloud. Selecting a point x in the part point cloudiCalculating said point xiDistance to each face element in the part model. And storing the surface elements corresponding to each distance into the stack in sequence according to the distances in ascending order. Find out the point xiProjection point x on top surface element of pilei' and determining the projection point xi' relationship to the stack top surface element boundaries. And repeating the steps until all points in the part point cloud are traversed. And dividing the points corresponding to the same surface element into the same point set according to the established corresponding relation. A set of points constitutes a segmented point cloud.
Description
Technical Field
The invention relates to the field of precision measurement, in particular to a part point cloud segmentation method based on geometric factors.
Background
The point cloud is a collection of point data on the surface of an object, and is a common model representation method. The point cloud contains information of all surfaces of the model, and is widely applied to the fields of error analysis, free-form surface reconstruction, machine vision and the like. In performing the correlation analysis of the model, it is the specific surface of the model that is targeted. Therefore, in the analysis, a subset of point clouds belonging to different patches needs to be segmented from the point clouds. The essence of point cloud segmentation is to divide points corresponding to different features into the same set of points based on the feature information of the model.
Currently, point cloud segmentation methods can be classified into three categories according to different processing modes: edge-based methods, face-based methods, and cluster-based methods. However, most edge-based methods are not only affected by measurement noise, but also are not easy to find the boundary of a curved surface with little curvature variation. When the surface-based method is used, the point cloud edge is easy to deform, and the problems of over-segmentation and under-segmentation occur. When using a clustering-based approach, the basis for the quasi-clustering may be found. Meanwhile, in clustering, a large amount of neighborhood information of points needs to be calculated. Therefore, the clustering-based method is easy to perform error segmentation on sparse point clouds or point clouds with uneven density.
The method is based on the local information of the point cloud when the point cloud data extracts the characteristic information of the model, thereby realizing the segmentation of the point cloud. Therefore, the above method is less accurate.
Disclosure of Invention
The present invention is directed to solving the problems of the prior art.
The technical scheme adopted for achieving the purpose of the invention is that the part point cloud segmentation method based on the geometric elements mainly comprises the following steps:
1) determining parts needing point cloud segmentation;
2) extracting geometric elements of the part;
further, the geometric elements include points, lines, and faces of the part.
3) Establishing a part model in an IGES format according to the geometric elements;
4) inputting the part model in Auto Cad;
5) scanning the part model, and extracting surface elements of the part model;
6) point sampling is carried out on the part model;
7) establishing a part point cloud according to the points sampled in the step 6;
8) carrying out coarse registration on the part model and the part point cloud by utilizing a directional bounding box method;
9) after coarse registration, performing fine registration on the part model and the part point cloud by using an iterative closest point algorithm;
10) after fine registration, selecting a point x in the point cloud of the parti;1≤i≤n
Wherein x isiAny point in the part point cloud; i is the serial number of the midpoint of the point cloud of the part, and the initial value of i is 1; n is the total number of the midpoint of the part point cloud;
11) calculating the point x according to the fine registration result obtained in step 10iDistance to each face element in the part model;
12) arranging the distances obtained by calculation in an ascending order; storing the distances in ascending order into a container Vector of VC + +;
13) according to the distances in ascending order, storing the surface elements corresponding to each distance into a stack in sequence; the initial face element of the top of the pile being x from said pointiSurface element P with minimum distance1(ii) a The initial face element of the pile bottom is x from the pointiThe surface element P with the largest distancem;
The set of face elements is P ═ { P ═ P1,P2,P3...Pm};
Wherein, P1,P2,P3...PmAre all surface elements; m is the total number of the face elements;
14) find out the point xiProjection point x on top surface element of pilei'; judging the projection point xi' relationship to the stack top surface element boundaries; if the projection point xi' inside the stack top surface element boundary, the point x is establishediCorresponding relation with the top surface element of the pile; if the projection point xiIf the element is outside the boundary of the top element of the pile, deleting the top element of the pile and updating the pile;
15) repeating step 14 until the corresponding relation is established or the surface elements are judged completely;
16) repeating steps 11-15 until all points in the part point cloud are traversed;
17) dividing points corresponding to the same surface element into the same point set according to the established corresponding relation; a set of points constitutes a segmented point cloud.
The technical effect of the present invention is undoubted. Aiming at the point cloud data of the parts, particularly the sparse point cloud data and the point cloud data with uneven density, the invention effectively combines the correlation algorithm of IGES model reconstruction and point cloud registration, and realizes the point cloud segmentation by utilizing the geometric elements of the IGES model. The invention effectively reduces the requirement of prior knowledge and improves the automation degree and precision of point cloud segmentation. Therefore, the invention can realize high-quality segmentation of the part point cloud.
Drawings
FIG. 1 is a flow chart of the implementation.
Fig. 2 is an origin cloud model.
Fig. 3 is a model after segmentation using the present invention.
Detailed Description
The present invention is further illustrated by the following examples, but it should not be construed that the scope of the above-described subject matter is limited to the following examples. Various substitutions and alterations can be made without departing from the technical idea of the invention and the scope of the invention is covered by the present invention according to the common technical knowledge and the conventional means in the field.
Example 1:
referring to fig. 1, a method for segmenting a point cloud of a part based on geometric elements mainly includes the following steps:
1) determining parts needing point cloud segmentation;
2) extracting geometric elements of the part;
further, the geometric elements include points, lines, and faces of the part. The method mainly utilizes the surface elements of the parts.
Preferably, when the geometric elements of the part are extracted, the surface of the part real object can be scanned and measured three-dimensionally from different positions and angles by using laser displacement, and the key parts can be measured in a thinning manner.
3) Establishing a part model in an IGES format according to the geometric elements;
further, the IGES format is mainly used for storing the geometric dimensions of the part model and the topological relationship of the point line surface. The IGES format is a NURBS-based file format that can be opened by Auto Cad. The part model can provide accurate part feature information, so that the time for feature extraction is saved, and the accuracy and the automation degree of point cloud segmentation are improved.
4) Inputting the part model in Auto Cad;
5) scanning the part model, and extracting surface elements of the part model;
6) point sampling is carried out on the part model;
preferably, in the point sampling process, some key parts are finely acquired, more sampling points are obtained as much as possible in a region with large curvature or severe curvature change, and less sampling points can be obtained in a relatively flat region.
7) Establishing a part point cloud according to the points sampled in the step 6;
further, the part point cloud is in an STL format; the STL file has a simple format and is mainly used for describing geometric information of a three-dimensional object. The part point cloud may be a sparse point cloud or a point cloud of non-uniform density.
8) Coarsely registering the part model and the part point cloud using an Oriented Bounding Box (OBB) method;
further, the directed bounding box approach replaces the complex geometric objects approximately with somewhat bulky and simple-featured geometries (bounding boxes). In selecting the size and orientation of the bounding box, it is based on the geometry of the part itself. That is, there is an arbitrary direction toward the bounding box. Therefore, the enclosing box can enclose the part as much as possible according to the shape characteristics of the part.
9) After coarse registration, performing fine registration on the part model and the part point cloud by using an iterative closest point algorithm (ICP);
further, the iterative closest point algorithm is mainly used for seeking a matching relation between the part model and the part point cloud.
10) After fine registration, selecting a point x in the point cloud of the parti;1≤i≤n
Wherein x isiAny point in the part point cloud; i is the serial number of the midpoint of the point cloud of the part, and the initial value of i is 1; n is the total number of the midpoint of the part point cloud;
11) calculating the point x according to the fine registration result obtained in step 10iDistance to each face element in the part model;
12) arranging the distances obtained by calculation in an ascending order; storing the distances in ascending order into a container Vector of VC + +;
13) according to the distances in ascending order, storing the surface elements corresponding to each distance into a stack in sequence; the initial face element of the top of the pile being x from said pointiSurface element P with minimum distance1(ii) a The initial face element of the pile bottom is x from the pointiThe surface element P with the largest distancem;
The set of face elements is P ═ { P ═ P1,P2,P3...Pm};
Wherein, P1,P2,P3...PmAre all surface elements; m is the total number of the face elements;
14) find out the point xiProjection point x on top surface element of pilei'; judging the projection point xi' relationship to the stack top surface element boundaries; if the projection point xi' inside the stack top surface element boundary, the point x is establishediCorresponding relation with the top surface element of the pile; if the projection point xiIf the element is outside the boundary of the top element of the pile, deleting the top element of the pile and updating the pile;
further, judging the projection point xi' the method of the boundary relation with the stack top surface element is: passing through the point xiAnd said projection point xi' make an arbitrary plane, and take an intersection of the plane and the stack top element boundary. The intersection point and the projection point xi' form a curved line segment. If there are an odd number of intersections of the curved line segment and the stack top surface element, the projection point xi' inside the stack top element boundaries. If there are an even number of intersections of the curved line segment and the stack top surface element, then the projection point xi' outside the stack top element boundaries.
15) Repeating step 14 until the corresponding relation is established or the surface elements are judged completely;
16) repeating steps 11 to 14 until all points in the part point cloud are traversed;
17) dividing points corresponding to the same surface element into a point set according to the established corresponding relation; a set of points constitutes a segmented point cloud.
Further, referring to fig. 2 and 3, the part point cloud segmentation method based on geometric elements segments a sparse point cloud or a point cloud with uneven density. The finally obtained segmentation point cloud ensures that the points in the same point set are consistent with the geometric element characteristics of the part as far as possible.
Claims (2)
1. A part point cloud segmentation method based on geometric elements is characterized by mainly comprising the following steps:
1) determining parts needing point cloud segmentation;
2) extracting geometric elements of the part;
3) establishing a part model in an IGES format according to the geometric elements;
4) inputting the part model in Auto Cad;
5) scanning the part model, and extracting surface elements of the part model;
6) point sampling is carried out on the part model;
7) establishing a part point cloud according to the points sampled in the step 6;
8) carrying out coarse registration on the part model and the part point cloud by utilizing a directional bounding box method;
9) after coarse registration, performing fine registration on the part model and the part point cloud by using an iterative closest point algorithm;
10) after fine registration, selecting a point x in the point cloud of the parti;1≤i≤n
Wherein x isiAny point in the part point cloud; i is the serial number of the midpoint of the point cloud of the part, and the initial value of i is 1; n is the total number of the midpoint of the part point cloud;
11) calculating the point x according to the fine registration result obtained in step 10iDistance to each face element in the part model;
12) arranging the distances obtained by calculation in an ascending order; storing the distances in ascending order into a container Vector of VC + +;
13) according to the distances in ascending order, storing the surface elements corresponding to each distance into a stack in sequence; the initial face element of the top of the pile being x from said pointiSurface element P with minimum distance1(ii) a The initial face element of the pile bottom is x from the pointiThe surface element P with the largest distancem;
The set of face elements is P ═ { P ═ P1,P2,P3...Pm};
Wherein, P1,P2,P3...PmAre all surface elements; m is the total number of the face elements;
14) find out the point xiProjection point x on top surface element of pilei'; judging the projection point xi' relationship to the stack top surface element boundaries; if the projection point xi' inside the stack top surface element boundary, the point x is establishediCorresponding relation with the top surface element of the pile; if the projection point xiIf the element is outside the boundary of the top element of the pile, deleting the top element of the pile and updating the pile;
15) repeating the step 14 until the corresponding relation is established or all the surface elements in the pile are judged;
16) repeating steps 11-15 until all points in the part point cloud are traversed;
17) dividing points corresponding to the same surface element into the same point set according to the established corresponding relation; a set of points constitutes a segmented point cloud.
2. The method for point cloud segmentation of parts based on geometric elements as claimed in claim 1, wherein: the geometric elements include points, lines and faces of the part.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710852830.0A CN107784656B (en) | 2017-09-20 | 2017-09-20 | Part point cloud segmentation method based on geometric elements |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710852830.0A CN107784656B (en) | 2017-09-20 | 2017-09-20 | Part point cloud segmentation method based on geometric elements |
Publications (2)
Publication Number | Publication Date |
---|---|
CN107784656A CN107784656A (en) | 2018-03-09 |
CN107784656B true CN107784656B (en) | 2020-02-07 |
Family
ID=61437647
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710852830.0A Active CN107784656B (en) | 2017-09-20 | 2017-09-20 | Part point cloud segmentation method based on geometric elements |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107784656B (en) |
Families Citing this family (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108986149A (en) * | 2018-07-16 | 2018-12-11 | 武汉惟景三维科技有限公司 | A kind of point cloud Precision Registration based on adaptive threshold |
CN112989453B (en) * | 2021-01-28 | 2022-12-06 | 徐洋洋 | BIM-based holographic deformation information extraction method |
CN113658184B (en) * | 2021-06-02 | 2024-03-12 | 重庆大学 | Point cloud curved surface segmentation method based on IGES model guidance |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102436654A (en) * | 2011-09-02 | 2012-05-02 | 清华大学 | Adaptive segmentation method of building point cloud |
CN102798362A (en) * | 2012-06-20 | 2012-11-28 | 北京航空航天大学 | Point cloud data-based method for estimating working allowance of casting |
CN103035006A (en) * | 2012-12-14 | 2013-04-10 | 南京大学 | High-resolution aerial image partition method based on LEGION and under assisting of LiDAR |
CN105844629A (en) * | 2016-03-21 | 2016-08-10 | 河南理工大学 | Automatic segmentation method for point cloud of facade of large scene city building |
EP3073443A1 (en) * | 2015-03-23 | 2016-09-28 | Université de Mons | 3D Saliency map |
-
2017
- 2017-09-20 CN CN201710852830.0A patent/CN107784656B/en active Active
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102436654A (en) * | 2011-09-02 | 2012-05-02 | 清华大学 | Adaptive segmentation method of building point cloud |
CN102798362A (en) * | 2012-06-20 | 2012-11-28 | 北京航空航天大学 | Point cloud data-based method for estimating working allowance of casting |
CN103035006A (en) * | 2012-12-14 | 2013-04-10 | 南京大学 | High-resolution aerial image partition method based on LEGION and under assisting of LiDAR |
EP3073443A1 (en) * | 2015-03-23 | 2016-09-28 | Université de Mons | 3D Saliency map |
CN105844629A (en) * | 2016-03-21 | 2016-08-10 | 河南理工大学 | Automatic segmentation method for point cloud of facade of large scene city building |
Also Published As
Publication number | Publication date |
---|---|
CN107784656A (en) | 2018-03-09 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Tazir et al. | CICP: Cluster Iterative Closest Point for sparse–dense point cloud registration | |
CN102750537B (en) | Automatic registering method of high accuracy images | |
CN107784656B (en) | Part point cloud segmentation method based on geometric elements | |
CN113178009B (en) | Indoor three-dimensional reconstruction method utilizing point cloud segmentation and grid repair | |
CN108830888B (en) | Coarse matching method based on improved multi-scale covariance matrix characteristic descriptor | |
CN103714541A (en) | Method for identifying and positioning building through mountain body contour area constraint | |
CN108303037A (en) | A kind of detection method and device of the workpiece surface shape difference based on cloud analysis | |
CN110634131B (en) | Crack image identification and modeling method | |
CN111179321A (en) | Point cloud registration method based on template matching | |
CN111783722B (en) | Lane line extraction method of laser point cloud and electronic equipment | |
CN111915657A (en) | Point cloud registration method and device, electronic equipment and storage medium | |
CN107316328A (en) | A kind of closed loop detection method based on two dimensional laser scanning instrument Corner Feature | |
Huang et al. | Registration algorithm for point cloud based on normalized cross-correlation | |
Yogeswaran et al. | 3d surface analysis for automated detection of deformations on automotive body panels | |
CN106886988B (en) | Linear target detection method and system based on unmanned aerial vehicle remote sensing | |
CN114463396B (en) | Point cloud registration method utilizing plane shape and topological graph voting | |
Kampel et al. | 3D puzzling of archeological fragments | |
US9443312B2 (en) | Line parametric object estimation | |
Yuan et al. | 3D point cloud recognition of substation equipment based on plane detection | |
Tong et al. | 3D point cloud initial registration using surface curvature and SURF matching | |
CN112581511B (en) | Three-dimensional reconstruction method and system based on near vertical scanning point cloud rapid registration | |
CN112435336B (en) | Curve type identification method and device, electronic equipment and storage medium | |
CN113536959A (en) | Dynamic obstacle detection method based on stereoscopic vision | |
Omidalizarandi et al. | Segmentation and classification of point clouds from dense aerial image matching | |
CN109670557B (en) | Automatic highway point cloud registration method based on rod-shaped structures |
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 | ||
GR01 | Patent grant | ||
GR01 | Patent grant |