CN111508073B - Method for extracting roof contour line of three-dimensional building model - Google Patents
Method for extracting roof contour line of three-dimensional building model Download PDFInfo
- Publication number
- CN111508073B CN111508073B CN202010168973.1A CN202010168973A CN111508073B CN 111508073 B CN111508073 B CN 111508073B CN 202010168973 A CN202010168973 A CN 202010168973A CN 111508073 B CN111508073 B CN 111508073B
- Authority
- CN
- China
- Prior art keywords
- line
- point
- contour line
- line segment
- new
- 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
- G06T17/00—Three dimensional [3D] modelling, e.g. data description of 3D objects
- G06T17/20—Finite element generation, e.g. wire-frame surface description, tesselation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2200/00—Indexing scheme for image data processing or generation, in general
- G06T2200/04—Indexing scheme for image data processing or generation, in general involving 3D image data
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2210/00—Indexing scheme for image generation or computer graphics
- G06T2210/04—Architectural design, interior design
Landscapes
- Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Computer Graphics (AREA)
- Geometry (AREA)
- Software Systems (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Image Analysis (AREA)
Abstract
A method for extracting a roof contour line of a three-dimensional building model comprises the following steps: step 1, preprocessing a three-dimensional building model, and extracting triangular patches belonging to a roof part; step 2, carrying out contour line rough extraction by adopting an Alpha Shapes algorithm; step 3, simplifying the rough contour line by a contour line simplifying method of a least squares method; and 4, regularizing the simplified contour line by adopting a classification forced orthogonal method. The invention uses Alpha Shapes algorithm and contour line simplification algorithm based on least square to obtain contour line with complete and accurate boundary; the clear and correct contour is ensured by adjusting the threshold value of the algorithm; and a classification forced regularization algorithm is utilized to obtain a regularized building outline, so that a basic method is provided for further processing of the three-dimensional building.
Description
Technical Field
The invention relates to the field of computer graphics, in particular to a method for extracting a roof contour line of a three-dimensional building model by utilizing a simple ring concept design contour search algorithm, which can obtain a building contour line with better integrity and higher geometric precision, and then, simplify the extracted initial contour by utilizing a contour line simplification algorithm with the minimum square; and finally, regularizing the contour line by using a classification forced orthogonal regularization algorithm.
Background
The contour information of the building is an important basis for building extraction and three-dimensional model reconstruction, and is widely applied to the fields of target identification, disaster estimation, real estate and the like. At present, the general method for tracking the plane contour of a building is to interpolate point clouds into a depth image, then segment the depth image by using an image segmentation algorithm, and finally track the boundary of the building by using methods such as a scanning line method and a neighborhood search method. The problem with these methods is that the traced edge is a rough boundary of the set of discrete points, with low accuracy. Some scholars also research methods for directly extracting outlines of discrete point sets, such as yellow pioneer and the like, and propose an edge tracking algorithm based on plane discrete points, wherein the algorithm takes a side length ratio as a constraint condition, and reduces the dependence of algorithm parameters on point density, so that the adaptability of the algorithm to extraction of edges of point sets with slender features or uneven distribution is improved, but the phenomenon of excessive shrinkage of the edges is easily caused by improper threshold setting of the constraint condition.
Extracting the building outline from the point cloud can be summarized to restore the original shape of the discrete point set, and the Alpha Shapes algorithm is an effective method. The Alpha Shapes algorithm is a deterministic algorithm with strict mathematical definition, and the shape of the point set obtained by the Alpha Shapes algorithm is determined for any finite point set S, and in addition, the user can control the fineness degree of the shape of the point set by adjusting a parameter Alpha unique to the algorithm. However, in practical applications, there are still some problems to be solved: the Alpha Shapes algorithm has a poor processing effect when processing the concave point cloud sets, if the Alpha value is large, the concave corners are easy to be passivated, if the Alpha value is small, the broken point set shape is easy to obtain, the Alpha value is set to be 1-2 times of the average point distance, and at the moment, the obtained point set shape is relatively complete and is not too broken.
Aiming at the problems, the invention provides a concept design contour search algorithm of a simple ring, which firstly uses an Alpha Shapes algorithm to completely extract the building contour; then, simplifying the extracted initial contour line by adopting a contour line simplification method based on least square; and finally, utilizing a classification forced orthogonal regularization algorithm to regularize the contour line.
Disclosure of Invention
The invention provides a method for extracting a three-dimensional building model roof contour line, which aims to overcome the defects of the prior art in extracting the three-dimensional building roof contour line, can quickly and accurately extract the contour line of any three-dimensional building, and simplifies and regularizes the extracted thick contour line segment, so that the method is better applied to actual production.
A method for extracting a roof contour line of a three-dimensional building model comprises the following specific steps:
step 1, preprocessing a three-dimensional building model, and extracting triangular patches belonging to a roof part;
step 2, carrying out contour line rough extraction by adopting an Alpha Shapes algorithm;
step 3, simplifying the rough contour line by a contour line simplification method based on least square;
and 4, regularizing the simplified contour line by adopting a classification forced orthogonal method.
The technical conception of the invention is as follows: extracting a rough closed contour line of the building by using an Alpha Shapes algorithm; further simplifying the contour line by using a least square contour line simplification method; and utilizing a classification forced orthogonal regularization algorithm to regularize the contour line. Compared with the classic Douglas Peucker algorithm, the contour line simplified by the method is closer to the actual contour line under the condition of noise.
The invention has the advantages that: obtaining a contour line with complete and accurate boundary by utilizing an Alpha Shapes algorithm and a contour line simplification algorithm based on least square; the clear and correct outline is ensured by adjusting the threshold value of the algorithm; and a classification forced regularization algorithm is utilized to obtain a regularized building outline, so that a basic method is provided for further processing of the three-dimensional building.
Drawings
FIG. 1 is a general flow diagram of the present invention.
FIG. 2 is an original three-dimensional building model.
Fig. 3 is a schematic diagram of extraction of a roof panel.
FIG. 4 is a schematic diagram of the Alpha Shapes algorithm after extracting the rough contour edge.
FIG. 5 is a diagram illustrating a simplified method based on least squares after simplifying contour edges.
FIG. 6 is a schematic diagram after regularizing the contour edges.
Detailed Description
The invention is further illustrated with reference to the accompanying drawings:
the method for extracting the roof contour line of the three-dimensional building model comprises the following steps:
step 1, preprocessing a three-dimensional building model, and extracting triangular patches belonging to a roof part;
step 2, carrying out contour line rough extraction by adopting an Alpha Shapes algorithm;
step 3, simplifying the rough contour line by a contour line simplification method based on least square;
and 4, regularizing the simplified contour line by adopting a classification forced orthogonal method.
In the step 1, aiming at a single three-dimensional building model, the invention firstly extracts a triangular patch of a roof surface. Specifically, a triangular patch with a patch height larger than h and a patch normal vector and a ground normal vector included angle smaller than theta is regarded as a roof patch. h can be freely adjusted, and the value of 1/2 of the height of the building is generally taken; θ took 20 degrees in all experiments. Fig. 2 shows the original three-dimensional building model to be simplified according to the present invention, and fig. 3 shows green areas for preprocessing extracted roof patches.
In the step 2, after the roof surface pieces are extracted, the outer contour lines of the roof surface are extracted. The method for extracting the roof outer contour line through the adjacency relation of the triangular surface patches is a common triangular mesh contour line extraction method, but the method is easy to misjudge the broken inner contour as the outer contour. The invention adopts Alpha Shapes method to solve the problem, the method is an effective contour line extraction algorithm for recovering the original shape of the discrete point set, and the extraction process is equal to rolling around the point set by using a circle with radius Alpha. Specifically, starting from any point P in all vertex sets S of the triangular patch of the roof, a subset S is formed by points with a distance smaller than 2 alpha 2 Taking any point Q in the middle, and calculating the center of the P, Q pointO; then judging whether other points exist in the circle with the point O as the center and the radius of alpha, if not, P, Q is a boundary point, the line PQ is a boundary line segment, otherwise, the point set S is pointed to 2 Repeating the judgment at the next point until S 2 Finishing the judgment of all the points; and finally, repeating the judgment by taking the next point in the point set S until the judgment of all the points in S is finished. When the above-mentioned flow is finished, all boundary line segments are extracted, and connected in sequence so as to obtain the rough contour line.
In step 3, the contour line segment extracted by the Alpha Shapes method is very rough (FIG. 4), and generally needs to be further simplified. The invention provides a contour line simplification method based on a least squares method, and the detailed design steps are as follows. The method requires two parameters, namely a maximum distance threshold d and a minimum straight line length threshold l from a point to a straight line, and the regularized result is shown in fig. 5. In the present invention, d is generally set to 0.5 times the average dot pitch, and l is set to 3 times the average dot pitch. In order to arrange in order outer contour two-dimensional point set U s The method takes a simplified two-dimensional point set which is arranged in sequence as input and output, and comprises the following specific steps:
(3.1) sequentially selecting the U on the contour line s Fitting a straight line L passing through the three vertexes by using a least squares method, calculating the distance from the three vertexes to the straight line L, and if the distance from any one vertex to the straight line L is greater than a threshold value d, turning to the 4 th step; otherwise, let the set U = { a, b, c };
(3.2) taking two end points of the set U as p and q, respectively extending towards two sides by taking the p and the q as starting points, encountering a new vertex in the extending process, adding the new vertex into the set U if the distance from the new vertex to the straight line L is less than a threshold value d, and extending again by taking the new vertex as a new starting point; otherwise stopping the extension at the vertex;
(3.3) in the step 2, if a new vertex is added into the set U, refitting a straight line L to all the vertices in the set U by using a least squares method, and turning to the step 2; otherwise, turning to the step 4;
(3.4) judging the length of the straight line fitted by the set U, if the length of the straight line is greater than a threshold value l, reserving vertexes at two ends of the set U, and abandoning a vertex in the middle; otherwise, discarding all vertexes;
and (3.5) if three continuous vertexes are not judged, turning to the step 1.
In said step 4, the building usually has a relatively regular geometric shape. Therefore, aiming at the extraction of the outer contour line of the building, the two most basic characteristics of vertical and parallel are fully considered, the method for classifying the forced orthogonal regularization contour line provided by the invention mainly classifies all the outer contour line segments uniformly according to direction vectors, and divides the outer contour line segments into two mutually orthogonal line segments, and the specific method is that the longest line segment is found out as an initial line segment, other line segments are judged with the initial line segment, if the difference of the azimuth angles of the two line segments is within the range of a threshold value, the line segment is classified as A, otherwise, the line segment is classified as B; then, each line segment is adjusted according to the length weighted average azimuth angle of each line segment, and a new end point of the outgoing line segment is calculated. The specific method is to keep the average azimuth angle of the A-type line segment unchanged, and force the average azimuth angle of the B-type line segment to be orthogonal with the A-type line segment to obtain a new average azimuth angle of the B-type line segment. And then adjusting each line segment, namely adjusting the line segment to a new average azimuth angle by taking the center point of the line segment as an axis, and solving the intersection point of two-by-two intersection of the new line segment as a new endpoint. Finally, all new vertices are connected end to end in sequence to obtain a regularized outer contour of the building, as shown in fig. 6.
At present, extracting contour lines from building point clouds is a popular research direction in data processing, and aiming at the problem that precision and integrity are difficult to be considered when the point cloud building contour is extracted, a concept design contour searching algorithm utilizing a simple ring is provided to obtain an initial building contour; then, a contour line simplification algorithm with the least square is used for simplifying the extracted initial contour line, compared with the classic Douglas Peucker algorithm, the contour line simplified by the method is closer to the actual contour line under the condition of noise; and finally, regularizing the contour by using a classification forced orthogonal method. The invention solves the problem of extracting the outline of the building and obtains the outline of the building with better integrity and higher geometric precision.
The embodiments described in this specification are merely illustrative of implementations of the inventive concept and the scope of the present invention should not be considered limited to the specific forms set forth in the embodiments but rather by the equivalents thereof as may occur to those skilled in the art upon consideration of the present inventive concept.
Claims (1)
1. A method for extracting a roof contour line of a three-dimensional building model comprises the following steps:
step 1, preprocessing a three-dimensional building model, and extracting triangular patches belonging to a roof part; the method specifically comprises the following steps: aiming at a single three-dimensional building model, firstly extracting a triangular patch of a roof surface, regarding the triangular patch with the patch height larger than h and the patch normal vector and ground normal vector included angle smaller than theta as the roof patch, and taking h as a value of 1/2 of the building height; taking theta as 20 degrees;
step 2, carrying out contour line rough extraction by adopting an Alpha Shapes algorithm; the method specifically comprises the following steps: extracting the outer contour line of the roof surface by using an Alpha Shapes method, wherein the extraction process is equal to rolling around a point set by using a circle with the radius of Alpha, starting from any point P in all vertex sets S of the triangular patch of the roof and forming a subset S by using points with the distance of less than 2 Alpha 2 Taking any point Q in the three points, and solving the circle center O of P, Q; then judging whether other points exist in a circle with the radius alpha and the point O as the center of the circle, if not, P, Q is a boundary point, a line PQ is a boundary line segment, otherwise, the point set S is checked 2 Repeating the judgment at the next point until S 2 Finishing the judgment of all the points; finally, the next point in the point set S is taken to repeat the judgment until all the points in S are judged to be finished; after the flow is finished, all boundary line segments are extracted and connected in sequence to obtain a rough contour line;
step 3, simplifying the rough contour line by a contour line simplifying method of a least squares method; the method specifically comprises the following steps: a maximum distance threshold d from the point to the straight line and a minimum straight line length threshold l, d is set to be 0.5 times of the average point distance, and l is set to be 3 times of the average point distance; in order of sequential arrangement of two-dimensional points of outer contourCollection U s The method takes a simplified two-dimensional point set which is arranged in sequence as input and output, and comprises the following specific steps:
(3.1) sequentially selecting the U on the contour line s Fitting a straight line L passing through the three vertexes by using a least squares method, calculating the distance from the three vertexes to the straight line L, and if the distance from any one vertex to the straight line L is greater than a threshold value d, turning to the 4 th step; otherwise, let the set U = { a, b, c };
(3.2) taking two end points of the set U as p and q, respectively extending towards two sides by taking the p and the q as starting points, encountering a new vertex in the extending process, adding the new vertex into the set U if the distance from the new vertex to the straight line L is less than a threshold value d, and extending again by taking the new vertex as a new starting point; otherwise stopping extension at the vertex;
(3.3) in the step (3.2), if a new vertex is added into the set U, refitting a straight line L by using a least squares method for all the vertices in the set U, and turning to the step (3.2); otherwise, turning to the step (3.4);
(3.4) judging the length of the straight line fitted by the set U, if the length of the straight line is greater than a threshold value l, reserving vertexes at two ends of the set U, and abandoning a vertex in the middle; otherwise, discarding all vertexes;
(3.5) if three continuous vertexes are not judged, turning to the step (3.1);
step 4, regularizing the simplified contour line by adopting a classification forced orthogonal method; the method specifically comprises the following steps: all the outer contour line segments are uniformly classified according to the direction vectors and are divided into two types of mutually orthogonal line segments, and the specific method comprises the following steps: finding out that the longest line segment is the initial line segment, judging other line segments with the initial line segment, if the difference of the azimuth angles of the two line segments is within the threshold range, classifying the line segment as A, and otherwise classifying the line segment as B; then, adjusting each line segment according to the length weighted average azimuth angle of each line segment, and calculating a new end point of the outgoing line segment; keeping the average azimuth angle of the A-type line segment unchanged, and forcibly orthogonal the average azimuth angle of the B-type line segment to obtain a new average azimuth angle of the B-type line segment; then adjusting each line segment, namely adjusting the line segment to a new average azimuth angle by taking the center point of the line segment as an axis, and solving the intersection point of two-by-two intersection of the new line segment as a new end point; and finally, sequentially connecting all the new vertexes end to obtain a regularized building outer contour line.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202010168973.1A CN111508073B (en) | 2020-03-12 | 2020-03-12 | Method for extracting roof contour line of three-dimensional building model |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202010168973.1A CN111508073B (en) | 2020-03-12 | 2020-03-12 | Method for extracting roof contour line of three-dimensional building model |
Publications (2)
Publication Number | Publication Date |
---|---|
CN111508073A CN111508073A (en) | 2020-08-07 |
CN111508073B true CN111508073B (en) | 2023-03-31 |
Family
ID=71877636
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202010168973.1A Active CN111508073B (en) | 2020-03-12 | 2020-03-12 | Method for extracting roof contour line of three-dimensional building model |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN111508073B (en) |
Families Citing this family (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112348836B (en) * | 2020-11-06 | 2024-03-12 | 二十一世纪空间技术应用股份有限公司 | Method and device for automatically extracting building outline |
CN112802041A (en) * | 2021-02-05 | 2021-05-14 | 北京工业大学 | Ground laser point cloud building contour line extraction method based on improved Alpha Shapes algorithm |
CN112819106B (en) * | 2021-04-16 | 2021-07-13 | 江西博微新技术有限公司 | IFC component type identification method, device, storage medium and equipment |
CN113496050B (en) * | 2021-05-27 | 2024-02-13 | 中机中联工程有限公司 | Logistics building roof comprehensive design method combining CAD and intelligent algorithm |
CN113298838B (en) * | 2021-07-27 | 2021-09-21 | 成都睿铂科技有限责任公司 | Object contour line extraction method and system |
CN114636403A (en) * | 2022-02-21 | 2022-06-17 | 同济大学 | Power transmission tower inclination measuring method based on three-dimensional laser point cloud |
CN115631307B (en) * | 2022-11-17 | 2023-03-21 | 北京飞渡科技有限公司 | Building segmented contour extraction and vertical face three-dimensional reconstruction method |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2013101428A (en) * | 2011-11-07 | 2013-05-23 | Pasuko:Kk | Building contour extraction device, building contour extraction method, and building contour extraction program |
CN106570468A (en) * | 2016-10-25 | 2017-04-19 | 中国人民解放军空军工程大学 | Method for reconstructing LiDAR original point cloud building contour |
CN109993783A (en) * | 2019-03-25 | 2019-07-09 | 北京航空航天大学 | A kind of roof and side optimized reconstruction method towards complex three-dimensional building object point cloud |
CN110827402A (en) * | 2020-01-13 | 2020-02-21 | 武大吉奥信息技术有限公司 | Method and system for simplifying three-dimensional model of similar building based on rasterization technology |
Family Cites Families (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US9189862B2 (en) * | 2010-06-10 | 2015-11-17 | Autodesk, Inc. | Outline approximation for point cloud of building |
US9286717B2 (en) * | 2013-07-30 | 2016-03-15 | Hewlett-Packard Development Company, L.P. | 3D modeling motion parameters |
-
2020
- 2020-03-12 CN CN202010168973.1A patent/CN111508073B/en active Active
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2013101428A (en) * | 2011-11-07 | 2013-05-23 | Pasuko:Kk | Building contour extraction device, building contour extraction method, and building contour extraction program |
CN106570468A (en) * | 2016-10-25 | 2017-04-19 | 中国人民解放军空军工程大学 | Method for reconstructing LiDAR original point cloud building contour |
CN109993783A (en) * | 2019-03-25 | 2019-07-09 | 北京航空航天大学 | A kind of roof and side optimized reconstruction method towards complex three-dimensional building object point cloud |
CN110827402A (en) * | 2020-01-13 | 2020-02-21 | 武大吉奥信息技术有限公司 | Method and system for simplifying three-dimensional model of similar building based on rasterization technology |
Non-Patent Citations (2)
Title |
---|
"Identification of nonlinear dynamic load models using an alternating least squares approach";Jing Yang;《2011 4th International Conference on Electric Utility Deregulation and Restructuring and Power Technologies (DRPT)》;第1155-1160页 * |
"基于激光点云数据的建筑物快速三维建模";饶杰;《中国优秀硕士学位论文全文数据库 (信息科技辑)》(第1期);第I138-752页 * |
Also Published As
Publication number | Publication date |
---|---|
CN111508073A (en) | 2020-08-07 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN111508073B (en) | Method for extracting roof contour line of three-dimensional building model | |
CN106780524B (en) | Automatic extraction method for three-dimensional point cloud road boundary | |
CN103727930B (en) | A kind of laser range finder based on edge matching and camera relative pose scaling method | |
CN101847163B (en) | Design patent image retrieval method with multi-characteristics fusion | |
CN109685080B (en) | Multi-scale plane extraction method based on Hough transformation and region growth | |
CN105678747B (en) | A kind of tooth mesh model automatic division method based on principal curvatures | |
CN109615581B (en) | Splicing recovery method of three-dimensional fragments fusing expanded Gaussian balls and color geometric features | |
CN111145228A (en) | Heterogeneous image registration method based on local contour point and shape feature fusion | |
CN102938161B (en) | A kind of 3D shape automatic division method based on Mean Shift | |
CN110807781A (en) | Point cloud simplification method capable of retaining details and boundary features | |
CN106600617A (en) | Method of extracting building contour line from Lidar point cloud data based on curvature | |
CN111354083B (en) | Progressive building extraction method based on original laser point cloud | |
CN113628263A (en) | Point cloud registration method based on local curvature and neighbor characteristics thereof | |
CN113409332B (en) | Building plane segmentation method based on three-dimensional point cloud | |
CN107545602B (en) | Building modeling method under space topological relation constraint based on LiDAR point cloud | |
CN108428236B (en) | Multi-target SAR image segmentation method based on feature fair integration | |
CN112802041A (en) | Ground laser point cloud building contour line extraction method based on improved Alpha Shapes algorithm | |
Liu et al. | A novel rock-mass point cloud registration method based on feature line extraction and feature point matching | |
CN107784646A (en) | A kind of road self-adapting detecting method to gather materials | |
CN111639560A (en) | Finger vein feature extraction method and device based on dynamic fusion of vein skeleton line and topographic relief characteristic | |
CN114387329B (en) | Building contour progressive regularization method based on high-resolution remote sensing image | |
CN117132630A (en) | Point cloud registration method based on second-order spatial compatibility measurement | |
CN109508674B (en) | Airborne downward-looking heterogeneous image matching method based on region division | |
CN112767429A (en) | Ground-snow surface point cloud rapid segmentation method | |
CN104036515A (en) | Vehicle target height estimation method based on single SAR image |
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 |