CN118097069B - Vegetation filtering method and device for mountain terrain grid model and computer equipment - Google Patents
Vegetation filtering method and device for mountain terrain grid model and computer equipment Download PDFInfo
- Publication number
- CN118097069B CN118097069B CN202410527826.7A CN202410527826A CN118097069B CN 118097069 B CN118097069 B CN 118097069B CN 202410527826 A CN202410527826 A CN 202410527826A CN 118097069 B CN118097069 B CN 118097069B
- Authority
- CN
- China
- Prior art keywords
- point set
- grid
- vegetation
- points
- point
- 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
- 238000000034 method Methods 0.000 title claims abstract description 56
- 238000001914 filtration Methods 0.000 title claims abstract description 26
- 239000013598 vector Substances 0.000 claims description 16
- 238000012216 screening Methods 0.000 claims description 11
- 238000004040 coloring Methods 0.000 claims description 9
- 238000004422 calculation algorithm Methods 0.000 claims description 8
- 238000004590 computer program Methods 0.000 claims description 5
- 230000001172 regenerating effect Effects 0.000 claims description 4
- 239000011435 rock Substances 0.000 claims description 2
- 238000000513 principal component analysis Methods 0.000 claims 1
- 230000010365 information processing Effects 0.000 abstract description 2
- 238000010586 diagram Methods 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 2
- 238000006467 substitution reaction Methods 0.000 description 2
- 238000012876 topography Methods 0.000 description 2
- 240000003243 Thuja occidentalis Species 0.000 description 1
- 235000008109 Thuja occidentalis Nutrition 0.000 description 1
- 230000003044 adaptive effect Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000012937 correction Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 230000004927 fusion Effects 0.000 description 1
- 238000005286 illumination Methods 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 230000000877 morphologic effect Effects 0.000 description 1
- 238000012847 principal component analysis method Methods 0.000 description 1
- 230000000717 retained effect Effects 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Abstract
The invention provides a vegetation filtering method and device of a mountain terrain grid model and computer equipment, and belongs to the technical field of information processing. The vegetation filtering method comprises the steps of firstly obtaining three-dimensional grid information calculated by three-dimensional reconstruction software, then extracting a regional point set T1 which does not need to be processed, fitting a fitting plane W for triangulation, obtaining point sets T2, T3 and T4 according to Delaunay triangulation, obtaining a ground surface point P 3D in the point set T3 after the vegetation characteristics are displayed, and repeating the steps until a vegetation point P 3Z in the point set T3 is an empty set; and obtaining the mountain terrain grid model with filtered vegetation after recovering the topological relation of the point set T4. The invention combines three-dimensional reconstruction software and the existing delaunay triangulation method, and abandons a complex region growing method, and the method can accurately model and easily filter vegetation in a mountain terrain grid model.
Description
Technical Field
The invention belongs to the technical field of information processing, and particularly relates to a vegetation filtering method and device of a mountain terrain grid model and computer equipment.
Background
In the field of engineering, the survey work on the terrain at the beginning of the engineering is often of great guiding significance to the design and construction of the whole engineering. The unmanned aerial vehicle photogrammetry technology greatly facilitates the topographic survey work, has the characteristics of low operation and maintenance cost and higher flexibility, but the topographic data acquired by unmanned aerial vehicle photogrammetry often comprises a large amount of surface vegetation information, and the vegetation information is not considered in engineering generally. Therefore, there is a need in the art for a method of rapidly filtering vegetation that maximally characterizes the true relief of the earth while maintaining valid earth data points and yet filters vegetation as a great difficulty in topography measurements.
Chinese patent application CN202310501303.0 discloses a method for removing vegetation in high-gradient areas. Comprising the following steps: projecting all points in the point set P to be processed onto an xoy plane, and performing 2D Delaunay triangulation to obtain a triangulation set T; calculating the gradient and the slope direction of each point in the point set P; classifying the point set P according to the gradient and the slope direction; finding a seed triangle tseed in the triangle net set T; gradually changing the z coordinate of the vegetation area point by using an area growth method; for points where the z coordinate values have changed, and points where the z coordinate values of adjacent points have changed, they are referred to as untrusted points; deleting the unreliable points to obtain a point set P' after vegetation is removed. The method can obtain more accurate results by carrying out the earthwork calculation, thereby solving the difficult problem of cost control; the accuracy rate of landslide prediction is higher, and the loss of agriculture caused by landslide can be better prevented.
However, the inventors of the present application have found that while the reconstructed mesh is considered in this patent application to be real surface information, in practice it will generate a malformed mesh and thus its constructed model does not conform to real surface conditions.
Accordingly, there remains a need in the art for a new vegetation filtering method, apparatus and computer device for a mountain land grid model.
Disclosure of Invention
The invention firstly provides a vegetation filtering method of a mountain terrain grid model, which comprises the following steps:
S1, acquiring three-dimensional grid information calculated by three-dimensional reconstruction software, wherein the three-dimensional grid information comprises integral point set information P, grid topology information F and coloring information;
S2, extracting a regional point set T1 which does not need to be processed through coloring information and manual screening;
S3, fitting a fitting plane W for triangulation according to the integral point set information P, traversing all grid surfaces of the grid topology information F, extracting points which do not accord with Delaunay triangulation to obtain a point set T2, wherein the point set T2 is a point set formed by points in a suspected vegetation area;
S4, firstly finding out points which do not accord with Delaunay triangulation in the point set T1 to obtain a point set T11, and then removing all points in the point set T11 from the point set T2 to obtain a point set T3 to be changed, wherein the point set T3 is the point set formed by points in a vegetation area; the whole point set information P comprises a ground surface point P D and other points, and the point set T3 consists of a vegetation point P 3Z and a ground surface point P 3D; the whole point set information P is also composed of a point set T3 to be changed and a point set T4 not to be changed;
S5, reconstructing the topological relation of the integral point set information P by adopting Delaunay triangulation according to the fitting plane W, so that vegetation characteristics are visualized, and new grid topological information is obtained;
S6, according to the new grid topology information, obtaining normal vector information of the grid where each point is located, wherein the normal vector information is gradient information, screening earth surface points P D according to the normal vector information, reserving an intersection part between the point set T3 and the earth surface points P D, namely reserving earth surface points P 3D in the point set T3, and deleting vegetation points P 3Z in the point set T3;
s7, repeating the steps S5-S6, regenerating grids and denoising until the conditions are met, namely, the vegetation points P 3Z in the point set T3 are deleted in a number of times, and the vegetation points P 3Z are empty sets;
s8, correcting the grid regenerated in the step S7, and recovering the topological relation of the unchanged point set T4; and obtaining the mountain terrain grid model with filtered vegetation.
In a specific embodiment, in step S1, the three-dimensional reconstruction software is METASHAPE or Autodesk ReCap.
In a specific embodiment, in step S2, the set of points T1 includes a house, a rock wall, a lake and a river.
In a specific embodiment, in step S3, a fitting plane W for triangulation is fitted using a plane fitting algorithm comprising at least one of a least squares method, a principal component analysis method and a RANSAC method.
In a specific embodiment, in step S3, the entire point set information P, i.e., the point set P, is projected onto the fitting plane W to obtain the point set p_w; and in step S5, the whole grid point set information P is split by using a Bowyer-Watson algorithm, and the algorithm can generate a triangular grid surface by calculating the distance between each point and other points.
In one embodiment, step S5 comprises in particular the steps of,
S5.1: acquiring a point set P_W and a fitting plane W according to the step S3;
S5.2: selecting three points in a point set P in a W plane to form a large triangle, wherein the circumscribed circle of the large triangle is required to contain all the points;
s5.3: adding points one by one, and for each newly added point, finding all triangles whose circumscribed circles contain the point, wherein the triangles do not meet Delaunay properties and therefore need to be removed;
s5.4: deleting all triangles of the circumscribed circle found in the previous step, which contain the new addition point, thereby generating a polygonal hole in triangulation; then connecting the new points with boundary points of the polygonal holes to form a series of new triangles, namely new grid surfaces;
S5.5: repeating S5.3 and S5.4 until all points are added to the triangle subdivision; after the triangulation is performed on the point set p_w, a triangle with the triangulation meeting the criterion Delaunay is finally obtained, that is, the circumscribed circle of each triangle does not contain any point, and at this time, the corresponding new mesh topology information f_w is:
，
Wherein f i is the grid surface in the new grid, N_W is the total number of grid surfaces of the new grid, and f N_W is the last grid surface in the new grid;
s5.6: the points in the point set p_w are re-projected back to the point set P, i.e. the original spatial position of the point set P is restored.
The invention also provides a vegetation filtering device of the mountain terrain grid model, which is a computer readable storage medium, and a computer program is stored on the vegetation filtering device, and the computer program realizes each step of the vegetation filtering method of the mountain terrain grid model when being executed by a processor.
The invention also provides a computer device comprising a processor and a memory storing a computer program which, when executed by the processor, implements the steps of a vegetation filtering method of a mountain terrain grid model as described above.
Compared with the prior art, the invention has at least the following advantages: 1. in the prior patent application of the background technology, the point cloud data is directly processed, and a Delaunay triangulation technology is adopted to generate grids; the grid generated by the method may not accurately reflect the real topographic features; because the topological relation between the point clouds is independent of visual information, the topological relation is constructed only based on geometric distance; the invention has the function of more accurate three-dimensional reconstructed grid, and can maintain a better grid structure for a specific area. 2. The plane of delaunay D subdivision in the invention is a fitting plane W conforming to the gradient of the area, rather than a unified X-Y horizontal plane, and the invention can be better used for screening vegetation areas. 3. The core of the method is a grid result generated by delaunay triangulation, and the method is different from a high-order grid generation mode such as the distribution of the arborvitae in three-dimensional reconstruction software, the method adopts delaunay subdivision, and the topological relation among points is disturbed, so that an actually nonexistent malformed grid appears as shown in fig. 5, but the method can display the characteristics of the ground surface points, at the moment, the gradient information of the grid points is not an actual gradient, but the gradient characteristics of the ground surface points are quite obvious, and the screening of the ground surface points is very easy under the condition. Therefore, although the present invention and the prior patent application in the background art both use delaunay triangulation to generate grids, the present invention can clearly filter vegetation without using the complex region growing method in the prior patent application, and can not construct a mountain land topography grid model which does not meet the real surface conditions. That is, the invention combines three-dimensional reconstruction software and delaunay triangulation method in the prior patent application, and abandons the complex region growing method, so that the method is an establishment method capable of easily filtering vegetation and accurately forming a mountain terrain grid model.
In general, the invention adopts a fusion topological relation, and the grid characteristics of a non-vegetation area are not destroyed when vegetation is filtered. The triangulation of the invention adopts an adaptive method to generate the fitting plane W, which is more beneficial to vegetation screening. The invention utilizes the morphological characteristics of vegetation to display the characteristics of the ground points covered by the vegetation under the condition of carrying out triangular sectioning on the plane delaunay again. Compared with other methods, the method has the advantages of small calculated amount and obvious characteristics.
Drawings
Fig. 1 is a schematic structural diagram of a fitting plane W.
Fig. 2 is a projection view of a grid structure in three-dimensional reconstruction software on a fitting plane W.
Fig. 3 is a three-dimensional perspective view of a mesh structure in three-dimensional reconstruction software.
Fig. 4 is a projection view of the mesh structure on the fitting plane W after mesh reconstruction using Delaunay triangulation.
FIG. 5 is a three-dimensional perspective view of a mesh structure after mesh reconstruction using Delaunay triangulation.
Fig. 6 is a schematic diagram of screening of the surface point P 3D.
FIG. 7 is a flow chart of a vegetation filtering method of the mountain terrain mesh model of the present invention.
In the figure, 1 is a point in the whole point set information P, 2 is a terrain grid, 3 is a fitting plane W,4 is a point in the vegetation point P 3Z, 5 is a point in the point set T4, and 6 is a point in the surface point P 3D.
Detailed Description
Step S1: and acquiring three-dimensional grid information calculated by the existing three-dimensional reconstruction software, wherein the three-dimensional grid information comprises integral point set information P, grid topology information F and coloring information.
The specific process comprises the following steps: the method comprises the steps of obtaining a grid M and extracting 1 of the grid file obtained through METASHAPE or Autodesk ReCap and other three-dimensional reconstruction software, wherein the whole point set information P is as follows:
，
For the following
，
Wherein: for coordinate information in xyz three directions of the point P i, N is the total number of point-concentrated points.
And simultaneously acquiring grid topology information F of a grid surface, wherein the grid topology information F is as follows:
，
Wherein:
，
Is three points constituting the surface f i, and
，
M is the total number of grid planes.
Coloring information: for a point cloud P, the point cloud P is:
，
obtaining RGB color information for each point, i.e
、、。
Step S2: the regional point set T1 which does not need to be processed is extracted through coloring information and manual screening.
The specific process comprises the following steps: setting a vegetation RGB threshold Range according to vegetation type, illumination conditions, camera settings and other conditions, and taking a red channel Range (R) in general: [0,100]; green channel Range (G): [80,255]; blue channel Range (B): [0,100]. And for the point cloud P, judging whether the color of the point is within a threshold range in sequence, and reserving points which are not within the range as terrain points to be reserved. By the following conditions:
points that do not meet the condition are retained, i.e., considered to be ground points, and noted as point cloud P M.
Meanwhile, non-vegetation coverage area point cloud information P N is extracted through manual judgment, and a union is taken as an area point set T1 which does not need to be processed, namely:
。
Step S3: fitting a fitting plane W for triangulation according to the whole point set information P, traversing all grid surfaces of the grid topology information F, extracting points which do not accord with Delaunay triangulation to obtain a point set T2, wherein the point set T2 is a point set formed by points in a suspected vegetation area.
The specific process comprises the following steps: for the point set P, adopting a least square method to perform plane fitting to obtain a fitting plane W, wherein the specific parameter expression is as follows:
Ax+By+Cz+D=0。
The set of points P is then projected onto the fitting plane W, resulting in a set of points p_w. And for P_W, constructing a grid by adopting the original topological relation, traversing each grid surface in F, judging whether the circumcircle of the triangular surface contains other grid points, if so, considering that the grid surface does not accord with Delaunay triangulation, and incorporating three points on the grid surface into a point set T2 formed by points in a suspected vegetation zone.
Step S4: firstly, finding out points which do not accord with Delaunay triangulation in a point set T1 to obtain a point set T11, and then removing all points in the point set T11 from a point set T2 to obtain a point set T3 to be changed, wherein the point set T3 is the point set formed by points in a vegetation area; the whole point set information P comprises a ground surface point P D and other points, and the point set T3 consists of a vegetation point P 3Z and a ground surface point P 3D; the whole point set information P is also composed of a point set T3 to be changed and a point set T4 not to be changed.
Step S5: and reconstructing the topological relation of the integral point set information P by adopting Delaunay triangulation according to the fitting plane W, thereby displaying vegetation characteristics and obtaining new grid topological information.
The method comprises the following specific steps: for the grid integral point set information P, a Bowyer-Watson algorithm is adopted for subdivision, and the algorithm can generate triangles by calculating the distance between each point and other points; the specific flow is as follows:
s5.1: acquiring a point set P_W and a fitting plane W according to the step S3;
S5.2: selecting three points in a point set P in a W plane to form a large triangle, wherein the circumscribed circle of the large triangle is required to contain all the points;
s5.3: adding points one by one, and for each newly added point, finding all triangles whose circumscribed circles contain the point, wherein the triangles do not meet Delaunay properties and therefore need to be removed;
s5.4: deleting all triangles of the circumscribed circle found in the previous step, which contain the new addition point, thereby generating a polygonal hole in triangulation; then, connecting the new points with boundary points of the polygonal holes to form a series of new triangles, namely new grid surfaces;
S5.5: repeating S5.3 and S5.4 until all points are added to the triangle subdivision; after the triangulation is performed on the point set p_w, a triangle with the triangulation meeting the criterion Delaunay is finally obtained, that is, the circumscribed circle of each triangle does not contain any point, and at this time, the corresponding new mesh topology information f_w is:
，
Wherein f i is the grid surface in the new grid, N_W is the total number of grid surfaces of the new grid, and f N_W is the last grid surface in the new grid;
s5.6: the points in the point set p_w are re-projected back to the point set P, i.e. the original spatial position of the point set P is restored.
Fig. 2 to 5 are corresponding drawings in step S5. Wherein, fig. 2 is a projection diagram of a grid structure in three-dimensional reconstruction software in a fitting plane W; fig. 3 is a three-dimensional perspective view of a mesh structure in three-dimensional reconstruction software. FIG. 4 is a projection view of a mesh structure on a fitting plane W after mesh reconstruction using Delaunay triangulation; FIG. 5 is a three-dimensional perspective view of a mesh structure after mesh reconstruction using Delaunay triangulation.
The principle of the step S5 of the invention is as follows: the screen triangulation is carried out on the space model, the low-point characteristics can be amplified, the terrain characteristic points of the vegetation lower part after the triangulation can be pointed, and the specific principle is as follows: for the slope body and mountain area containing vegetation, the vegetation growth angle and the surface fluctuation are always smaller than 90 degrees, meanwhile, the tree structure is always thinner, and the upper grid points are distributed more, so that the upper grid points are projected on the W surface, and the upper grid points of the tree are always covered with the terrain points. Therefore, the vegetation grid point cloud and the terrain grid information are projected to the W plane, delaunay triangulation is carried out again, the original terrain vegetation grid structure can be destroyed, corner features can exist at the intersection of vegetation and the ground, and the ground surface features of the vegetation bottom can be visualized through the mode.
Step S6: and acquiring normal vector information of the grid where each point is located according to the new grid topology information, wherein the normal vector information is gradient information, screening the ground surface points P D according to the normal vector information, reserving an intersection part between the point T3 and P D, namely reserving the ground surface points P 3D in the point set T3, and deleting the vegetation points P 3Z in the point set T3.
The specific process comprises the following steps: for new grid data, the whole point set information P, namely the point cloud P and the grid topology information F_w, is acquired. Step S5 indicates that the terrain points covered by vegetation form sharp-angled grids with the adjacent higher vegetation points after the grids are subdivided, so that it is only necessary to determine whether the new grid points are located in the grids with such sharp-angled grids, and then determine whether the points are the terrain points covered by vegetation.
The judging process is as follows:
S6.1: all mesh plane information f i0 including the point P i is acquired:
，
wherein,
；
S6.2: calculating unit normal vector of each grid surface in f i0 ：
Center point coordinates：
，
The center point coordinate calculation formula is as follows:
If the grid is known as:
，
The center point of the substitution grid is as follows:
；
S6.3: obtaining unit normal vector of grid where P i is located ：
，
And calculates an average vector：
Then calculate each normal vector and average vectorVariance of included angleAs an index of flatness of the grid points, varianceThe method comprises the following steps:
Wherein: Representing vectors And (3) withIs included in the plane of the first part; when the variance satisfiesJudging whether the point is on a malformed grid or not; wherein,As the case may be, 10 is generally taken;
S6.4: if the point is on the malformed grid, the ordinate through the point Relationship with the center point:
，
Judging whether the point is a ground surface point or not, and adding a point set P D;
Traversing all points in the point set P, judging whether the points are ground points or not, and obtaining a ground point set P D; editing the point set to be changed T3, reserving the surface points P 3D in the point set to be changed, and deleting the vegetation points P 3Z.
After step S6 is completed, the result shown in fig. 6 is obtained.
Step S7: repeating the steps S5-S6, regenerating the grid and denoising until the condition is met, namely deleting the vegetation point P 3Z in the point set T3 in a known manner, wherein the vegetation point P 3Z is an empty set.
The specific process comprises the following steps: because the terrain points obtained in the step S6 do not necessarily completely meet the requirements, the vegetation points and the terrain points still generate grids with deformity and sharp peaks due to the height difference problem, and therefore the evenness index of each grid point is passed againDenoising the newly generated grid, namely deleting allAfter the points of (a), the terrain mesh is regenerated until allIf the condition is satisfied, the vegetation point P 3Z in T3 is considered to be deleted by the sydney, and the step is terminated.
Step S8: correcting the grid regenerated in the step S7, and recovering the topological relation of the unchanged point set T4; and obtaining the mountain terrain grid model with filtered vegetation.
The method comprises the following specific steps: after step S7 is completed, a new mesh conforming to Delaunay subdivision is obtained, and the corresponding mesh topology is:
，
however, in the area where the point set T4 is not changed, the topological relation of the grid still depends on a simple geometric relation, so that the problems of grid dislocation, deformity and the like can be caused, and the grid is corrected. The step S8 specifically comprises the following steps:
s8.1: for original grid topology information F:
，
Traversing F, and reserving a grid surface which is completely formed by the elements of the point set T4, so as to obtain a grid surface set F_1, namely, three vertexes of any grid surface element in the F_1 are formed by the point elements in the point set T4;
S8.2: traversing all grid planes for a new grid topological structure F_w, deleting the grid planes as long as three vertexes in the grid planes contain T4 midpoint elements, and finally obtaining a modified grid plane set F_2;
s8.3: and combining the F_1 and the F_2 to obtain a new grid topological structure F_3, reconstructing the grid according to the F_3, filling the cavity in the grid to obtain a final grid, and finishing the correction of the grid regenerated in the step S7.
The flow chart according to fig. 7, the flow comprising the steps of:
S1, acquiring integral point set information P, grid topology information F and coloring information calculated by three-dimensional reconstruction software;
s2, extracting an area point set T1 which does not need to be processed in the whole point set information P through coloring information and manual screening;
s3, extracting a point set T2 in the suspected vegetation area in the whole point set information P through Delaunay triangulation characteristics;
s4, dividing the whole point set information P into a point set T3 to be changed and a point set T4 not to be changed through the information of the point sets T1 and T2;
s5, reconstructing the topological relation of the integral point set information P by adopting Delaunay triangulation, and visualizing vegetation characteristics;
s6, deleting vegetation points P 3Z in the point set T3 to be changed according to normal vector information;
s7, repeating the steps S5-S6, regenerating grids and denoising until the conditions are met;
and S8, correcting the mesh regenerated in the step S7 to obtain a mountain land terrain mesh model with filtered vegetation.
In general, the invention provides a vegetation filtering method, a vegetation filtering device and computer equipment of a mountain terrain grid model. The vegetation filtering method comprises the steps of firstly obtaining three-dimensional grid information calculated by three-dimensional reconstruction software, then extracting a regional point set T1 which does not need to be processed, fitting a fitting plane W for triangulation, obtaining point sets T2, T3 and T4 according to Delaunay triangulation, obtaining a ground surface point P 3D in the point set T3 after the vegetation characteristics are displayed, and repeating the steps until a vegetation point P 3Z in the point set T3 is an empty set; and obtaining the mountain terrain grid model with filtered vegetation after recovering the topological relation of the point set T4. The invention combines three-dimensional reconstruction software and the existing delaunay triangulation method, and abandons a complex region growing method, and the method can accurately model and easily filter vegetation in a mountain terrain grid model.
The foregoing is a further detailed description of the invention in connection with specific preferred embodiments, and is not intended to limit the practice of the invention to such description. It will be apparent to those skilled in the art that several simple deductions and substitutions can be made without departing from the spirit of the invention, and these are considered to be within the scope of the invention.
Claims (8)
1. A vegetation filtering method of a mountain terrain grid model is characterized by comprising the following steps:
S1, acquiring three-dimensional grid information calculated by three-dimensional reconstruction software, wherein the three-dimensional grid information comprises integral point set information P, grid topology information F and coloring information;
S2, extracting a regional point set T1 which does not need to be processed through coloring information and manual screening;
S3, fitting a fitting plane W for triangulation according to the integral point set information P, traversing all grid surfaces of the grid topology information F, extracting points which do not accord with Delaunay triangulation to obtain a point set T2, wherein the point set T2 is a point set formed by points in a suspected vegetation area;
S4, firstly finding out points which do not accord with Delaunay triangulation in the point set T1 to obtain a point set T11, and then removing all points in the point set T11 from the point set T2 to obtain a point set T3 to be changed, wherein the point set T3 is the point set formed by points in a vegetation area; the whole point set information P comprises a ground surface point P D and other points, and the point set T3 consists of a vegetation point P 3Z and a ground surface point P 3D; the whole point set information P is also composed of a point set T3 to be changed and a point set T4 not to be changed;
S5, reconstructing the topological relation of the integral point set information P by adopting Delaunay triangulation according to the fitting plane W, so that vegetation characteristics are visualized, and new grid topological information is obtained;
S6, according to the new grid topology information, obtaining normal vector information of the grid where each point is located, wherein the normal vector information is gradient information, screening earth surface points P D according to the normal vector information, reserving an intersection part between the point set T3 and the earth surface points P D, namely reserving earth surface points P 3D in the point set T3, and deleting vegetation points P 3Z in the point set T3;
s7, repeating the steps S5-S6, regenerating grids and denoising until the conditions are met, namely, the vegetation points P 3Z in the point set T3 are deleted in a number of times, and the vegetation points P 3Z are empty sets;
s8, correcting the grid regenerated in the step S7, and recovering the topological relation of the unchanged point set T4; and obtaining the mountain terrain grid model with filtered vegetation.
2. The method of vegetation filtering of a mountain terrain mesh model according to claim 1, wherein in step S1, the three-dimensional reconstruction software is METASHAPE or Autodesk ReCap.
3. A vegetation filtering method of a mountain terrain mesh model as claimed in claim 1, wherein in step S2, the point set T1 includes houses, rock walls, lakes and rivers.
4. A vegetation filtering method of a mountain terrain mesh model according to claim 1, characterized in that in step S3, a fitting plane W for triangulation is fitted using a plane fitting algorithm comprising at least one of least squares, principal component analysis and RANSAC.
5. The vegetation filtering method of the mountain terrain mesh model according to claim 1, wherein in step S3, the whole point set information P, i.e., the point set P, is projected to the fitting plane W to obtain the point set p_w; and in step S5, the whole grid point set information P is split by using a Bowyer-Watson algorithm, and the algorithm can generate a triangular grid surface by calculating the distance between each point and other points.
6. The method for filtering vegetation of a mountain terrain mesh model as claimed in claim 5, wherein the step S5 comprises the steps of,
S5.1: acquiring a point set P_W and a fitting plane W according to the step S3;
S5.2: selecting three points in a point set P in a W plane to form a large triangle, wherein the circumscribed circle of the large triangle is required to contain all the points;
s5.3: adding points one by one, and for each newly added point, finding all triangles whose circumscribed circles contain the point, wherein the triangles do not meet Delaunay properties and therefore need to be removed;
s5.4: deleting all triangles of the circumscribed circle found in the previous step, which contain the new addition point, thereby generating a polygonal hole in triangulation; then connecting the new points with boundary points of the polygonal holes to form a series of new triangles, namely new grid surfaces;
S5.5: repeating S5.3 and S5.4 until all points are added to the triangle subdivision; after the triangulation is performed on the point set p_w, a triangle with the triangulation meeting the criterion Delaunay is finally obtained, that is, the circumscribed circle of each triangle does not contain any point, and at this time, the corresponding new mesh topology information f_w is:
，
Wherein f i is the grid surface in the new grid, N_W is the total number of grid surfaces of the new grid, and f N_W is the last grid surface in the new grid;
s5.6: the points in the point set p_w are re-projected back to the point set P, i.e. the original spatial position of the point set P is restored.
7. A vegetation filtering apparatus of a mountain land grid model, characterized in that the apparatus is a computer-readable storage medium having stored thereon a computer program which, when executed by a processor, implements the steps of the vegetation filtering method of a mountain land grid model as claimed in any one of claims 1 to 6.
8. A computer device comprising a processor and a memory storing a computer program which, when executed by the processor, performs the steps of the vegetation filtering method of the mountain terrain mesh model of any of claims 1 to 6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410527826.7A CN118097069B (en) | 2024-04-29 | Vegetation filtering method and device for mountain terrain grid model and computer equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410527826.7A CN118097069B (en) | 2024-04-29 | Vegetation filtering method and device for mountain terrain grid model and computer equipment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN118097069A CN118097069A (en) | 2024-05-28 |
| CN118097069B true CN118097069B (en) | 2024-07-09 |
Family
ID=
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107507202A (en) * | 2017-09-28 | 2017-12-22 | 武汉大学 | A kind of vegetation rotary island towards high-resolution remote sensing image automates extracting method |
| CN110910435A (en) * | 2019-11-08 | 2020-03-24 | 国网通用航空有限公司 | Building point cloud extraction method and device, computer equipment and readable storage medium |
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107507202A (en) * | 2017-09-28 | 2017-12-22 | 武汉大学 | A kind of vegetation rotary island towards high-resolution remote sensing image automates extracting method |
| CN110910435A (en) * | 2019-11-08 | 2020-03-24 | 国网通用航空有限公司 | Building point cloud extraction method and device, computer equipment and readable storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105783810B (en) | Engineering earthwork measuring method based on unmanned plane camera work | |
| CN109523630B (en) | Contour line generation method and device based on tension spline interpolation | |
| CN109242862B (en) | Real-time digital surface model generation method | |
| CN106327577B (en) | Dimensional topography optimal method based on local curvature's entropy and quad-tree structure | |
| KR102415768B1 (en) | A color image generating device by feature ground height and a color image generating program by feature height | |
| Pepe et al. | Techniques, tools, platforms and algorithms in close range photogrammetry in building 3D model and 2D representation of objects and complex architectures | |
| KR102415767B1 (en) | Color image generating device by feature height and color image generating program by feature height | |
| CN111784840B (en) | LOD (line-of-sight) level three-dimensional data singulation method and system based on vector data automatic segmentation | |
| CN102930509A (en) | Intelligent filtering method for airborne laser point cloud data | |
| CN111667574B (en) | Method for automatically reconstructing regular facade three-dimensional model of building from oblique photography model | |
| CN111612903B (en) | Geological data visualization method based on mixed data model | |
| CN109636889B (en) | Large-scale three-dimensional terrain model rendering method based on dynamic sewing belt | |
| Cosso et al. | Surveying and mapping a cave using 3d laser scanner: the open challenge with free and open source software | |
| CN108198194A (en) | Oblique photography single object extraction method based on lateral multi-plane cutting | |
| CN115797288B (en) | Method for calculating filling and excavating amount based on ground point cloud data | |
| CN109242951A (en) | A kind of face's real-time three-dimensional method for reconstructing | |
| CN104751479A (en) | Building extraction method and device based on TIN data | |
| CN111445569B (en) | Sedimentary geological evolution dynamic simulation method | |
| CN112084548B (en) | 2.5D visual modeling method and system for slope stability analysis | |
| CN117058339A (en) | Construction progress quantitative tracking method based on AS-SGM-SF voxel three-dimensional reconstruction | |
| KR101021013B1 (en) | A system for generating 3-dimensional geographical information using intensive filtering an edge of building object and digital elevation value | |
| CN113932779A (en) | Earth and stone calculation method and system based on unmanned aerial vehicle oblique photography and storage medium | |
| CN118097069B (en) | Vegetation filtering method and device for mountain terrain grid model and computer equipment | |
| CN111028349B (en) | Hierarchical construction method suitable for rapid visualization of massive three-dimensional live-action data | |
| CN118097069A (en) | Vegetation filtering method and device for mountain terrain grid model and computer equipment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant |