CN104318622B - Triangular mesh modeling method of indoor scene inhomogeneous three dimension point cloud data - Google Patents
Triangular mesh modeling method of indoor scene inhomogeneous three dimension point cloud data Download PDFInfo
- Publication number
- CN104318622B CN104318622B CN201410581951.2A CN201410581951A CN104318622B CN 104318622 B CN104318622 B CN 104318622B CN 201410581951 A CN201410581951 A CN 201410581951A CN 104318622 B CN104318622 B CN 104318622B
- Authority
- CN
- China
- Prior art keywords
- point
- neighborhood
- dimensional
- delaunay
- cloud data
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T17/00—Three dimensional [3D] modelling, e.g. data description of 3D objects
- G06T17/30—Polynomial surface description
Abstract
The invention relates to a triangular mesh modeling method of indoor scene inhomogeneous three dimension point cloud data. The triangular mesh modeling method of the indoor scene inhomogeneous three dimension point cloud data includes following steps: step 1, obtaining the indoor scene inhomogeneous three dimension point cloud data; step 2, performing neighborhood tangent plane projection; step 3, optimizing a projection neighborhood N'; step 4, obtaining a Delaunay adjacent side of a given point p; step 5, completing triangular mesh modeling: repeating the steps from the step 2 to the step 4, repeating the above algorithm on all points, and then completing the triangular mesh modeling of a whole three dimension point cloud. The triangular mesh modeling method of the indoor scene inhomogeneous three dimension point cloud data uses uniform partition of a sector area and all direction equilibrium selection, selects adjacent points of the given point in all directions well, evenly distributes the neighborhood in all the directions, and achieves precise modeling of the indoor scene inhomogeneous three dimension point cloud data. Simultaneously, selection of nearest neighbor points in all the directions further effectively streamlines the neighborhood of the given point, and therefore the triangular mesh modeling method of the indoor scene inhomogeneous three dimension point cloud data is short in running time and high in modeling efficiency.
Description
Technical field
A kind of a kind of the present invention relates to triangle gridding modeling method, more particularly, it relates to indoor scene non-homogeneous three-dimensional
The triangle gridding modeling method of cloud data.
Background technology
With the development of 3-D scanning ranging technology, three dimensional point cloud is in reverse-engineering, industrial detection, independent navigation etc.
The application in field is more and more extensive.Three dimensional point cloud treatment technology, as the basis realizing above-mentioned application, has played to pass
Important effect.In three dimensional point cloud treatment technology, the modeling of the triangle gridding of three dimensional point cloud be one very crucial
Technology.Because indoor environment is a kind of structuring scene, therefore triangle gridding modeling technique is particularly suitable for the three of indoor scene
Dimension modeling, it not only can describe indoor scene in lifelike image ground, and lays for the classification and target recognition of indoor scene
Good basis.The introducing of outstanding triangle gridding modeling method can significantly improve practical situations, improves application performance.
When obtaining indoor scene three dimensional point cloud, the operating characteristic of Laser Distance Measuring Equipment progressive scan and indoor environment structure prominent
So change, easily cause the unstable of scanning line spacing, so that the distribution of three dimensional point cloud becomes extremely uneven, give
The triangle gridding modeling of indoor scene brings larger difficulty.The triangle gridding modeling of three dimensional point cloud is always three-dimensional point
The study hotspot of cloud data processing field, its modeling method is broadly divided into two big class:Modeling based on Delaunay trigonometric ratio
Method and the modeling method of region growth.In general, the modeling method based on Delaunay trigonometric ratio is although there is good fortune
Row result, but need substantial amounts of computing, so that the execution efficiency of its algorithm is relatively low, modeling speed is slower;The modeling that region increases
Method has good operational efficiency, and modeling speed is very fast, but its modeling effect is not satisfactory sometimes.It is directed to non-homogeneous three-dimensional
The triangle gridding modeling of cloud data, rare report both at home and abroad at present, existing triangle gridding modeling technique is no longer applicable.Example
As relatively more famous rolling ball algorithm (Ball-Pivoting Algorithm) is accomplished by repeatedly transporting according to different size of ball
Go and to process uneven three dimensional point cloud, and result is unsatisfactory sometimes;For another example, based on two-dimentional Delaunay trigonometric ratio
Curve reestablishing method, it lays particular emphasis on the structure of the sampled point Delaunay adjacent side in incisal plane, and just their backprojection
To in three dimensions, to form triangle grid model, the method speed of service is very fast, but sampling condition and point cloud distribution are had
Strict restriction, it is impossible to process non-homogeneous three-dimensional cloud data, in each local triangle, hardly results in accurately
Delaunay adjacent side, Holistic modeling effect is poor.
Content of the invention
In order to overcome the deficiencies in the prior art, it is an object of the present invention to provide a kind of indoor scene non-homogeneous three-dimensional point
The triangle gridding modeling method of cloud data.The method is directed to an indoor scene, obtains room first with laser scanning and ranging instrument
The non-homogeneous three-dimensional cloud data of interior scene, it is substantially a non-homogeneous point set in three dimensions, then passes through certain
This point set is configured to a triangular mesh topological structure by surface modeling methods, with the real indoor scene of accurate description.Should
Method solves that the modeling quality brought due to cloud data skewness is relatively low, cannot describe actual scene and true
The problems such as topological structure deviates from, and also there is modeling speed faster.
In order to realize foregoing invention purpose, in the presence of solving the problems, such as prior art, the technical scheme that the present invention takes
It is:A kind of triangle gridding modeling method of indoor scene non-homogeneous three-dimensional cloud data, comprises the following steps:
Step 1, acquisition indoor scene non-homogeneous three-dimensional cloud data:By laser sensor, obtain indoor scene information,
As non-homogeneous three-dimensional cloud data;
Step 2, the projection of neighborhood incisal plane:Choose set point p=(x, y, z), calculate the average distance d of three-dimensional point cloud, if
Determining 5d is the radius of neighbourhood, obtains this given neighborhood of a point N={ pi=(xi, yi, zi) | 1≤i≤k }, wherein:piFor adjoint point, i is
The sequence number of adjoint point, k is the number of adjoint point, by the neighborhood N around p point, calculates the normal vector n of this point p;By this normal vector n,
Build the incisal plane T at p point, and by adjoint point piProject on the T of incisal plane, the point set after note projection is projection neighborhood N ';
Step 3, to projection neighborhood N ' be optimized:It is evenly dividing with each to equilibrium selection by sector region, further
Optimize the projection neighborhood N ' of set point p, make set point p all have nearest projection adjoint point, the point after note optimization in all directions
Collect for optimizing neighborhood N ";
Step 4, the Delaunay adjacent side of acquisition set point p:Using two-dimentional Delaunay method to set point p and its optimization
Neighborhood N " carries out triangle gridding modeling, by the two-dimentional Delaunay triangulation network lattice back mapping obtaining to three dimensional neighborhood space, and
Extract the three-dimensional Delaunay adjacent side being connected with storage with set point p;
Step 5, complete triangle gridding modeling:Repeat step 2-4, repeats above-mentioned algorithm to each point, then completes whole
The triangle gridding modeling of three-dimensional point cloud.
The projection of described step 2 neighborhood incisal plane, specifically includes following sub-step:
Step (a), the average distance d of calculating three-dimensional point cloud, set 5d as the radius of neighbourhood, extract the neighborhood N of set point p;
Step (b), pass through formula
Ask for the covariance matrix M of neighborhood N, in formula:piFor adjoint point, i is the sequence number of adjoint point, and k is the number of adjoint point, and T is
Vectorial transposition symbol, column vector transposition is row vector by it;
Step (c), ask for the eigenvalue λ of M1、λ2、λ3(λ1<λ2<λ3), and corresponding characteristic vector v1、v2、v3;
Step (d), by minimal eigenvalue λ1Corresponding characteristic vector v1Unitization, that is, obtain the normal vector n of set point p;
On step (e), the incisal plane T that each point in neighborhood N is all projected corresponding to normal vector n, after projection
Point set is designated as projecting neighborhood N '.
Described step 3 is optimized to projection neighborhood N ', specifically includes following sub-step:
Step (a), in the T of incisal plane, with set point p as initial point, built the straight line of initial point p, and be allowed to from horizontal position
Put beginning dextrorotation to circle, at interval of 22.5 degree, one sector region of division, finally can form 16 with initial point p is
The sector region of the heart;
Step (b), the projection neighborhood N ' in p point incisal plane T is inserted ready-portioned sector region in step (a)
Interior;
Step (c), extract in each sector region apart from the point that initial point p is nearest, these points just constitute the point sets after optimizing,
It is designated as optimizing neighborhood N ".
Described step 4 obtains the Delaunay adjacent side of set point p point, specifically includes following sub-step:
Step (a), the method using dividing and ruling, will optimize neighborhood N " according to x coordinate, it is divided into several zonules, each
Point number in zonule is not more than 3, it is desirable to Delaunay criterion can be ensured compliance with for each zonule;
Step (b), neighbor cell domain is integrated into a larger region meeting Delaunay criterion;
Step (c), repeat step (b), then each large area is successively merged, until all large areas merge into one
Individual entirety, so far, the local Delaunay triangle division of single completes;According to the uniqueness criterion of Delaunay trigonometric ratio, can
To obtain unique Delaunay trigonometric ratio result;
Step (d), from build two-dimentional Delaunay triangle division result extract the side being connected with set point p, by it
Back mapping is to three dimensional neighborhood space, and is stored, stored while be set point p point three-dimensional Delaunay while.
Present invention has the advantages that:A kind of triangle gridding modeling method of indoor scene non-homogeneous three-dimensional cloud data, bag
Include following steps:Step 1, acquisition indoor scene non-homogeneous three-dimensional cloud data:By laser sensor, obtain indoor scene letter
Breath, as non-homogeneous three-dimensional cloud data;Step 2, the projection of neighborhood incisal plane:Choose set point p=(x, y, z), calculate three-dimensional
The average distance d of point cloud, sets 5d as the radius of neighbourhood, obtains this given neighborhood of a point N={ pi=(xi, yi, zi)|1≤i≤
K }, wherein:piFor adjoint point, i is the sequence number of adjoint point, and k is the number of adjoint point, by the neighborhood N around p point, calculates the method for this point p
Vector n;By this normal vector n, build the incisal plane T at p point, and by adjoint point piProject on the T of incisal plane, the point after note projection
Collect for projecting neighborhood N ';Step 3, to projection neighborhood N ' be optimized:It is evenly dividing with each to equilibrium selection by sector region,
Optimize the projection neighborhood N ' of set point p further, make set point p all have nearest projection adjoint point in all directions, note optimizes
Point set afterwards is to optimize neighborhood N ";Step 4, the Delaunay adjacent side of acquisition set point p:Using two-dimentional Delaunay method to giving
Fixed point p and its optimization neighborhood N " carries out triangle gridding modeling, by the two-dimentional Delaunay triangulation network lattice back mapping obtaining to three
Dimension neighborhood space, and extract and store the three-dimensional Delaunay adjacent side being connected with set point p;Step 5, complete triangle gridding and build
Mould:Repeat step 2-4, repeats above-mentioned algorithm to each point, then completes the triangle gridding modeling of whole three-dimensional point cloud.With
Technology is compared, and the present invention is evenly dividing with each to equilibrium selection using sector region, preferably have selected in all directions to
The adjoint point of fixed point, is allowed to neighborhood distribution in all directions and more equalizes it is achieved that indoor scene non-homogeneous three-dimensional point cloud number
According to Accurate Model.Meanwhile, respectively to the given neighborhood of a point of selection also effective Herba Stellariae Saxatilis letter of nearest neighbor point, this method is made to have
There is the modeling efficiency of lower run time and Geng Gao.
Brief description
Fig. 1 is the flow chart of the present invention.
Fig. 2 is the schematic diagram of step of the present invention.
In figure:A () is neighborhood incisal plane projection figure, (b) is to be optimized figure to projection neighborhood N ', and (c) is to obtain to give
The Delaunay adjacent side figure of point p, (d) is the Delaunay modeling figure under incisal plane.
Fig. 3 is the part plan triangle gridding modeling schematic diagram being provided without the present invention.
Fig. 4 is the triangle gridding modeling result figure of indoor scene 1 non-homogeneous three-dimensional cloud data.
In figure:A () is whole structure figure, (b), (c) are partial, detailed view respectively.
Fig. 5 is the triangle gridding modeling result figure of indoor scene 2 non-homogeneous three-dimensional cloud data;
In figure:A () is whole structure figure, (b), (c) are partial, detailed view respectively.
Specific embodiment
The invention will be further described below in conjunction with the accompanying drawings.
As shown in figure 1, a kind of triangle gridding modeling method of indoor scene non-homogeneous three-dimensional cloud data, walk including following
Suddenly:Step 1, acquisition indoor scene non-homogeneous three-dimensional cloud data:By laser sensor, obtain indoor scene information, as
Non-homogeneous three-dimensional cloud data;Step 2, the projection of neighborhood incisal plane:Choose set point p=(x, y, z), calculate the flat of three-dimensional point cloud
All apart from d, set 5d as the radius of neighbourhood, obtain this given neighborhood of a point N={ pi=(xi, yi, zi) | 1≤i≤k }, wherein:pi
For adjoint point, i is the sequence number of adjoint point, and k is the number of adjoint point, by the neighborhood N around p point, calculates the normal vector n of this point p;Pass through
This normal vector n, builds the incisal plane T at p point, and by adjoint point piProject on the T of incisal plane, the point set after note projection is that projection is adjacent
Domain N ';Step 3, to projection neighborhood N ' be optimized:It is evenly dividing with each to equilibrium selection by sector region, optimize further
The projection neighborhood N ' of set point p, makes set point p all have nearest projection adjoint point in all directions, the point set after note optimizes is
Optimize neighborhood N ";Step 4, the Delaunay adjacent side of acquisition set point p:Using two-dimentional Delaunay method to set point p and its
Optimize neighborhood N " carry out triangle gridding modeling, will be empty to three dimensional neighborhood for the two-dimentional Delaunay triangulation network lattice back mapping obtaining
Between, and extract and store the three-dimensional Delaunay adjacent side being connected with set point p;Step 5, complete triangle gridding modeling:Repeat to walk
Rapid 2-4, repeats above-mentioned algorithm to each point, then completes the triangle gridding modeling of whole three-dimensional point cloud.
The projection of described step 2 neighborhood incisal plane, specifically includes following sub-step:
Step (a), the average distance d of calculating three-dimensional point cloud, set 5d as the radius of neighbourhood, extract the neighborhood N of set point p;
Step (b), pass through formula
Ask for the covariance matrix M of neighborhood N, in formula:piFor adjoint point, i is the sequence number of adjoint point, and k is the number of adjoint point, and T is
Vectorial transposition symbol, column vector transposition is row vector by it;
Step (c), ask for the eigenvalue λ of M1、λ2、λ3(λ1<λ2<λ3), and corresponding characteristic vector v1、v2、v3;
Step (d), by minimal eigenvalue λ1Corresponding characteristic vector v1Unitization, that is, obtain the normal vector n of set point p;
On step (e), the incisal plane T that each point in neighborhood N is all projected corresponding to normal vector n, after projection
Point set is designated as projecting neighborhood N ';As shown in figure (a) in Fig. 2.
Described step 3 is optimized to projection neighborhood N ', specifically includes following sub-step:
Step (a), in the T of incisal plane, with set point p as initial point, built the straight line of initial point p, and be allowed to from horizontal position
Put beginning dextrorotation to circle, at interval of 22.5 degree, one sector region of division, finally can form 16 with initial point p is
The sector region of the heart;
Step (b), the projection neighborhood N ' in p point incisal plane T is inserted ready-portioned sector region in step (a)
Interior;
Step (c), extract in each sector region apart from the point that initial point p is nearest, these points just constitute the point sets after optimizing,
It is designated as optimizing neighborhood N ";As shown in figure (b) in Fig. 2.
Described step 4 obtains the Delaunay adjacent side of set point p point, specifically includes following sub-step:
Step (a), the method using dividing and ruling, will optimize neighborhood N " according to x coordinate, it is divided into several zonules, each
Point number in zonule is not more than 3, it is desirable to Delaunay criterion can be ensured compliance with for each zonule;
Step (b), neighbor cell domain is integrated into a larger region meeting Delaunay criterion;
Step (c), repeat step (b), then each large area is successively merged, until all large areas merge into one
Individual entirety, so far, the local Delaunay triangle division of single completes;According to the uniqueness criterion of Delaunay trigonometric ratio, can
To obtain unique Delaunay trigonometric ratio result;
Step (d), from build two-dimentional Delaunay triangle division result extract the side being connected with set point p, by it
Back mapping is to three dimensional neighborhood space, and is stored, stored while be set point p point three-dimensional Delaunay while;As
In Fig. 2 shown in figure (c).Rerun afterwards, repeat step 2-4 is put to each in original point cloud data, is finally cut flat with
Face modeling is as shown in figure (d) in Fig. 2.The side of all storages is shown according to trigonometric ratio result, you can obtain shown in Fig. 4 and Fig. 5
Triangle gridding modeling result.
In sum, the present invention does not carry out direct triangle gridding modeling to the discrete point in three-dimensional point cloud, but first right
It carries out the projection of neighborhood incisal plane, recycles sector region to divide to optimize the projection neighborhood of set point with each to equilibrium selection,
The distribution being allowed in all directions more equalizes, and final realizes local triangle grid using each to optimizing neighborhood in a balanced way
The foundation of model.This not only solves the accurate modeling problem of indoor scene non-homogeneous three-dimensional cloud data, it is to avoid modeling process
The generation in middle large area cavity, as shown in figure 3, truly describe the topological structure of indoor scene, and due to optimization process
Presence so that during local neighborhood modeling the quantity of discrete point drastically decline, it is to avoid the modeling of local domain mass data point and
The huge operation expense brought, drastically increases the operational efficiency of modeling, decreases the required run time of modeling.
Claims (4)
1. a kind of triangle gridding modeling method of indoor scene non-homogeneous three-dimensional cloud data is it is characterised in that include following walking
Suddenly:
Step 1, acquisition indoor scene non-homogeneous three-dimensional cloud data:By laser sensor, obtain indoor scene information, as
Non-homogeneous three-dimensional cloud data;
Step 2, the projection of neighborhood incisal plane:Choose set point p=(x, y, z), calculate the average distance d of three-dimensional point cloud, set 5d
For the radius of neighbourhood, obtain this given neighborhood of a point N={ pi=(xi,yi,zi) | 1≤i≤k }, wherein:piFor adjoint point, i is adjoint point
Sequence number, k be adjoint point number, by the neighborhood N around p point, calculate the normal vector n of this point p;By this normal vector n, build
Incisal plane T at p point, and by adjoint point piProject on the T of incisal plane, the point set after note projection is projection neighborhood N ';
Step 3, to projection neighborhood N ' be optimized:It is evenly dividing with each to equilibrium selection by sector region, optimize further
The projection neighborhood N ' of set point p, makes set point p all have nearest projection adjoint point in all directions, the point set after note optimizes is
Optimize neighborhood N ";
Step 4, the Delaunay adjacent side of acquisition set point p:To set point p and its optimize neighborhood using two-dimentional Delaunay method
N " carries out triangle gridding modeling, by the two-dimentional Delaunay triangulation network lattice back mapping obtaining to three dimensional neighborhood space, and extracts
The three-dimensional Delaunay adjacent side being connected with set point p with storage;
Step 5, complete triangle gridding modeling:Repeat step 2-4, repeats above-mentioned algorithm to each point, then completes whole three-dimensional
The triangle gridding modeling of point cloud.
2. a kind of triangle gridding modeling method of indoor scene non-homogeneous three-dimensional cloud data according to claim 1, it is special
Levy and be:The projection of described step 2 neighborhood incisal plane, specifically includes following sub-step:
Step (a), the average distance d of calculating three-dimensional point cloud, set 5d as the radius of neighbourhood, extract the neighborhood N of set point p;
Step (b), pass through formula
Ask for the covariance matrix M of neighborhood N, in formula:piFor adjoint point, i is the sequence number of adjoint point, and k is the number of adjoint point, and T is that vector turns
Set symbol, column vector transposition is row vector by it;
Step (c), ask for the eigenvalue λ of M1、λ2、λ3, wherein λ1<λ2<λ3, and corresponding characteristic vector v1、v2、v3;
Step (d), by minimal eigenvalue λ1Corresponding characteristic vector v1Unitization, that is, obtain the normal vector n of set point p;
On step (e), the incisal plane T all projecting corresponding to normal vector n by each point in neighborhood N, by the point set after projection
It is designated as projecting neighborhood N '.
3. a kind of triangle gridding modeling method of indoor scene non-homogeneous three-dimensional cloud data according to claim 1, it is special
Levy and be:Described step 3 is optimized to projection neighborhood N ', specifically includes following sub-step:
Step (a), in the T of incisal plane, with set point p as initial point, built the straight line of initial point p, and be allowed to open from horizontal level
Beginning dextrorotation is circled, and at interval of 22.5 degree, divides a sector region, finally can form 16 centered on initial point p
Sector region;
Step (b), the projection neighborhood N ' in p point incisal plane T is inserted in ready-portioned sector region in step (a);
Step (c), extract in each sector region apart from the point that initial point p is nearest, these points just constitute the point sets after optimizing, and are designated as
Optimize neighborhood N ".
4. a kind of triangle gridding modeling method of indoor scene non-homogeneous three-dimensional cloud data according to claim 1, it is special
Levy and be:Described step 4 obtains the Delaunay adjacent side of set point p point, specifically includes following sub-step:
Step (a), the method using dividing and ruling, will optimize neighborhood N and " according to x coordinate, be divided into several zonules, each cell
Point number in domain is not more than 3, it is desirable to Delaunay criterion can be ensured compliance with for each zonule;
Step (b), neighbor cell domain is integrated into a larger region meeting Delaunay criterion;
Step (c), repeat step (b), then each large area is successively merged, until all large areas merge into one whole
Body, so far, the local Delaunay triangle division of single completes;According to the uniqueness criterion of Delaunay trigonometric ratio, can obtain
To unique Delaunay trigonometric ratio result;
Step (d), from build two-dimentional Delaunay triangle division result extract the side being connected with set point p, it is reverse
Be mapped to three dimensional neighborhood space, and stored, stored while be set point p point three-dimensional Delaunay while.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201410581951.2A CN104318622B (en) | 2014-10-25 | 2014-10-25 | Triangular mesh modeling method of indoor scene inhomogeneous three dimension point cloud data |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201410581951.2A CN104318622B (en) | 2014-10-25 | 2014-10-25 | Triangular mesh modeling method of indoor scene inhomogeneous three dimension point cloud data |
Publications (2)
Publication Number | Publication Date |
---|---|
CN104318622A CN104318622A (en) | 2015-01-28 |
CN104318622B true CN104318622B (en) | 2017-02-22 |
Family
ID=52373848
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201410581951.2A Active CN104318622B (en) | 2014-10-25 | 2014-10-25 | Triangular mesh modeling method of indoor scene inhomogeneous three dimension point cloud data |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN104318622B (en) |
Families Citing this family (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104809759A (en) * | 2015-04-03 | 2015-07-29 | 哈尔滨工业大学深圳研究生院 | Large-area unstructured three-dimensional scene modeling method based on small unmanned helicopter |
CN106569270B (en) * | 2015-10-12 | 2018-10-02 | 中国石油化工股份有限公司 | The adaptive unstructured triangular grid method of regular grid rate pattern |
CN105608732B (en) * | 2015-12-18 | 2018-07-13 | 北京工业大学 | A kind of optimization method of triangle grid model |
CN109493380B (en) * | 2017-09-11 | 2022-05-17 | 重庆大学 | Method for calculating area of irregular shear surface in rock joint surface shear test |
CN108334080B (en) * | 2018-01-18 | 2021-01-05 | 大连理工大学 | Automatic virtual wall generation method for robot navigation |
CN108920574B (en) * | 2018-06-22 | 2022-09-30 | 黑龙江科技大学 | Efficient neighborhood searching method for large-scale three-dimensional point cloud data |
CN109671152B (en) * | 2018-12-01 | 2022-10-04 | 大连理工大学 | Line point cloud geometric characteristic estimation method based on Gaussian weighted discrete derivative |
US11816762B2 (en) * | 2019-05-27 | 2023-11-14 | Zhejiang University | Three-dimensional reconstruction method based on half-peak probability density distribution |
CN110264565B (en) * | 2019-05-27 | 2021-07-30 | 浙江大学 | Three-dimensional reconstruction method based on half-peak probability density distribution |
TWI766218B (en) | 2019-12-27 | 2022-06-01 | 財團法人工業技術研究院 | Reconstruction method, reconstruction system and computing device for three-dimensional plane |
CN111815691B (en) * | 2020-09-10 | 2020-12-18 | 熵智科技(深圳)有限公司 | Neighborhood access method, device, equipment and medium for grid data structure |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6968299B1 (en) * | 2000-04-14 | 2005-11-22 | International Business Machines Corporation | Method and apparatus for reconstructing a surface using a ball-pivoting algorithm |
CN1835021A (en) * | 2006-04-14 | 2006-09-20 | 东南大学 | Curve triangle section structure method of 3-D scattered point set in 3-D scanning system |
EP1793349A3 (en) * | 2005-12-05 | 2007-10-24 | Siemens Medical Solutions USA, Inc. | Method and aparatus for discrete mesh filleting and rounding through ball pivoting |
CN101377851A (en) * | 2007-08-29 | 2009-03-04 | 鸿富锦精密工业(深圳)有限公司 | System and method for computing a minimum distance from point cloud to point cloud |
CN102881047A (en) * | 2012-08-01 | 2013-01-16 | 桂林电子科技大学 | Automatic non-closed implicit curved surface reconstruction method |
-
2014
- 2014-10-25 CN CN201410581951.2A patent/CN104318622B/en active Active
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6968299B1 (en) * | 2000-04-14 | 2005-11-22 | International Business Machines Corporation | Method and apparatus for reconstructing a surface using a ball-pivoting algorithm |
EP1793349A3 (en) * | 2005-12-05 | 2007-10-24 | Siemens Medical Solutions USA, Inc. | Method and aparatus for discrete mesh filleting and rounding through ball pivoting |
CN1835021A (en) * | 2006-04-14 | 2006-09-20 | 东南大学 | Curve triangle section structure method of 3-D scattered point set in 3-D scanning system |
CN101377851A (en) * | 2007-08-29 | 2009-03-04 | 鸿富锦精密工业(深圳)有限公司 | System and method for computing a minimum distance from point cloud to point cloud |
CN102881047A (en) * | 2012-08-01 | 2013-01-16 | 桂林电子科技大学 | Automatic non-closed implicit curved surface reconstruction method |
Non-Patent Citations (5)
Title |
---|
3D mesh compression: survey, comparisons and emerging trends;ADRIEN MAGLO 等;《ACM Computing Surveys》;20130930;第39:1-39:40页 * |
An Analysis and Implementation of a Parallel Ball Pivoting Algorithm;Julie Digne;《Image Processing》;20140701;第149-168页 * |
The Ball-Pivoting Algorithm for Surface Reconstruction;Fausto Bernardini;《Visualization and Computer Graphics》;19991031;第349-359页 * |
一种快速三维散乱点云的三角剖分算法;陈伟 等;《计算机仿真》;20090930;第338-341页 * |
三维散乱点云快速曲面重建算法;宋大虎 等;《计算机应用研究》;20130831;第2527-2529页 * |
Also Published As
Publication number | Publication date |
---|---|
CN104318622A (en) | 2015-01-28 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN104318622B (en) | Triangular mesh modeling method of indoor scene inhomogeneous three dimension point cloud data | |
CN105654483B (en) | The full-automatic method for registering of three-dimensional point cloud | |
CN107862738B (en) | One kind carrying out doors structure three-dimensional rebuilding method based on mobile laser measurement point cloud | |
CN107392875A (en) | A kind of cloud data denoising method based on the division of k neighbours domain | |
CN103247041B (en) | A kind of dividing method of the cloud data of the many geometric properties based on local sampling | |
CN106127857B (en) | The on-board LiDAR data modeling method of integrated data driving and model-driven | |
CN103729872B (en) | A kind of some cloud Enhancement Method based on segmentation resampling and surface triangulation | |
CN107025685A (en) | Airborne building summit cloud modeling method under topology ambiguity | |
CN106780751A (en) | Three-dimensional point cloud method for reconstructing based on improved shielding Poisson algorithm | |
CN103701466A (en) | Scattered point cloud compression algorithm based on feature reservation | |
CN104504709B (en) | Feature ball based classifying method of three-dimensional point-cloud data of outdoor scene | |
CN105359163A (en) | Method for fitting primitive shapes to a set of 3D points | |
CN105761312A (en) | Micro-terrain surface reconstruction method | |
CN104361632A (en) | Triangular mesh hole-filling method based on Hermite radial basis function | |
CN104966317A (en) | Automatic three-dimensional modeling method based on contour line of ore body | |
CN106408608A (en) | Method for extracting trunk diameter from ground laser radar point cloud data | |
CN103559705B (en) | Computer method for comparing similarity of different plant forms | |
CN106886980A (en) | A kind of enhanced method of point cloud density based on three-dimensional laser radar target identification | |
CN109584357A (en) | Three-dimensional modeling method, device, system and storage medium based on more contour lines | |
CN103310481A (en) | Point cloud reduction method based on fuzzy entropy iteration | |
CN103106632A (en) | Fusion method of different-accuracy three-dimension point cloud data based on mean shift | |
CN105225272B (en) | A kind of tri-dimensional entity modelling method based on the reconstruct of more contour line triangulation networks | |
CN105069840A (en) | Three-dimensional normal distribution transformation point cloud registration method based on curvature feature | |
CN107038308B (en) | A kind of regular grid terrain modeling method based on linear interpolation | |
CN107610228A (en) | Curved surface increment topology rebuilding method based on massive point cloud |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
C14 | Grant of patent or utility model | ||
GR01 | Patent grant |