Method of supplementing detail texture for point cloud using images

By establishing Delaunay triangular network and collinear equations, computing and adding supplementary point cloud data, and giving true color information, the problem of sparse point cloud data is solved, and the effect of object recognition and classification is improved.

CN118644604BActive Publication Date: 2024-11-05CHENGDU HANGWEI ZHIXIN TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411112798.9
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-08-14
Publication Date
2024-11-05
Estimated Expiration
2044-08-14

AI Technical Summary

Technical Problem

Due to sparse collection of point cloud data, it is impossible to clearly display detailed textures in the visual interface, affecting object recognition and classification.

Method used

By obtaining point cloud data and its corresponding images, a Delaunay triangular network is established, a plane equation is fitted, a vertex is projected to the image plane, a supplementary point is selected and its coordinates and elevation in the point cloud coordinate system is calculated, and the supplementary point is added, and according to the Image RGB information gives true color information, enriching point cloud details texture.

Benefits of technology

It effectively supplements the detailed texture of point clouds, improves the density and visualization of point cloud data, and promotes object recognition and classification.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118644604B_ABST
    Figure CN118644604B_ABST
Patent Text Reader

Abstract

The present invention provides a method for supplementing detail texture for point cloud by using image, which relates to the technical field of image data processing. A Delaunay triangulation network is established according to point cloud data, and a sub-triangulation network is taken as an example to fit its plane equation to obtain the fitting plane equation, and the vertices of the sub-triangulation network are projected into the image by means of collinear equations to obtain the corresponding image plane triangulation area, and a supplementary point is selected in the image plane triangulation area, and the X-axis coordinate and the Y-axis coordinate in the point cloud coordinate system are inversely calculated by means of collinear equations, and then substituted into the fitting plane equation to solve the elevation, so as to add the supplementary point to the point cloud data, and according to the RGB information of the supplementary point in the image plane, the supplementary point is assigned true color information, so as to enrich the detail texture of the point cloud and solve the problem that the point cloud lacks detail texture. The present invention is suitable for supplementing the detail texture of the point cloud.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of image data processing, and in particular to a method for supplementing detail textures for point clouds using images. Background Art

[0002] Point cloud data records the real three-dimensional information of points on the surface of an object, especially the elevation information. However, since laser point cloud data is usually sparse during collection, after the point cloud is colored according to the image, the detailed texture cannot be clearly displayed in the point cloud data visualization interface, which is not conducive to object recognition and classification based on laser point cloud data.

[0003] The current common practice is to increase the density of point clouds during point cloud data collection to increase detail textures, but the cost is too high. Summary of the invention

[0004] Technical problem solved by the present invention: The present invention provides a method for supplementing detail texture for point cloud by using image, so as to solve the problem that existing point cloud lacks detail texture.

[0005] The present invention solves the above technical problems by adopting a technical solution: a method for supplementing detail textures for point clouds using images, comprising the following steps:

[0006] S1, obtain point cloud data and its corresponding images;

[0007] S2, establish a Delaunay triangulation network based on the point cloud data;

[0008] S3, selecting a sub-triangular network, fitting a fitting plane of the sub-triangular network, and obtaining a fitting plane equation;

[0009] S4, projecting the three vertices of the sub-triangle network onto the image plane respectively through collinear equations to obtain the image plane triangular region corresponding to the sub-triangle network;

[0010] S5, selecting a supplementary point in the triangulated area of ​​the image plane, calculating the X-axis coordinate and the Y-axis coordinate of the supplementary point in the point cloud coordinate system by using the collinear equation, substituting them into the fitting plane equation, calculating the elevation of the supplementary point, and adding the corresponding supplementary point to the point cloud data according to the X-axis coordinate, the Y-axis coordinate and the elevation;

[0011] S6, repeating S3 to S5, adding supplementary points to multiple sub-triangulated networks of the point cloud data;

[0012] S7. Assign true color information to the added supplementary points based on the RGB information of the supplementary points in the image plane.

[0013] Furthermore, the fitting plane equation is Z=AX+BY+C, where A, B and C are parameters, which are obtained by substituting the coordinate values ​​of the three vertices into the equation, or solved by the least squares method, Lagrange multiplier method or SVD decomposition method.

[0014] Furthermore, the collinear equation is: ,in,( , ) represents the coordinates of the vertex in the image plane, is the focal length of the camera, (X, Y, Z) represents the coordinates of the vertex in the point cloud coordinate system, ( , , ) represents the coordinates of the camera taking the image in the point cloud coordinate system, , , , , , , , and It is the direction cosines determined by the three angular elements of the image.

[0015] Further, in S5, the supplementary point includes the centroid of the triangular region or a series of points equidistant from the centroid of the triangular region.

[0016] Beneficial effects of the present invention: The present invention provides a method for supplementing detail texture for point clouds using images, establishing a Delaunay triangulation network based on point cloud data, taking one of the sub-triangulation networks as an example, fitting its plane equation to obtain the fitting plane equation, and projecting the vertices of the sub-triangulation network into the image through the collinear equation to obtain the corresponding image plane triangulation area, selecting supplementary points in the image plane triangulation area, using the collinear equation to inversely calculate its X-axis coordinate and Y-axis coordinate in the point cloud coordinate system, and then substituting them into the fitting plane equation to solve the elevation, thereby adding the supplementary points to the point cloud data, and assigning true color information to the supplementary points based on the RGB information of the supplementary points in the image plane, thereby enriching the detail texture of the point cloud, solving the problem of lack of detail texture in the point cloud, and facilitating subsequent object recognition and object classification based on laser point cloud data. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 It is a flow chart of a method for supplementing detail texture of point cloud using image provided by the present invention;

[0018] Figure 2 The present invention provides a Delaunay triangulation network established based on point cloud data. DETAILED DESCRIPTION

[0019] Aiming at the problem that point cloud data is sparse and cannot clearly display detailed textures, which is not conducive to object recognition and object classification based on laser point cloud data, the present invention provides a method for supplementing detailed textures for point clouds using images, such as Figure 1 As shown, the following steps are included:

[0020] S1, obtain point cloud data and its corresponding images;

[0021] S2, establish a Delaunay triangulation network based on point cloud data;

[0022] Specifically, the Delaunay triangulation network is established based on the point cloud data of a certain area, such as Figure 2 shown.

[0023] S3, selecting a sub-triangular network, fitting a fitting plane of the sub-triangular network, and obtaining a fitting plane equation;

[0024] Specifically, suppose the plane equation is: Z=AX+BY+C, the three vertices of the sub-triangle network are K1 (X1, Y1, Z1), K2 (X2, Y2, Z2) and K3 (X3, Y3, Z3), and the coordinates of the three vertices are substituted into the plane equation to obtain A, B and C. The plane equation can also be calculated using a computer program using the least squares method, Lagrange multiplier method or SVD decomposition method. The constraint condition used when using the Lagrange multiplier method or SVD decomposition method is that the sum of the squares of the distances from the three vertices to the fitting plane is minimized.

[0025] S4, projecting the three vertices of the sub-triangle network onto the image plane respectively through collinear equations to obtain the image plane triangular region corresponding to the sub-triangle network;

[0026] Specifically, the collinear equation is: ,in,( , ) represents the coordinates of the vertex in the image plane, is the focal length of the camera, (X, Y, Z) represents the coordinates of the vertex in the point cloud coordinate system, ( , , ) represents the coordinates of the camera taking the image in the point cloud coordinate system, , , , , , , , and It is the direction cosine determined by the three angular elements of the image. The three vertices of the sub-triangle network are projected onto the image plane through collinear equations to obtain the image plane coordinates of the three vertices. The triangle range formed by the image plane coordinates of the three vertices is the image plane triangle area corresponding to the sub-triangle network.

[0027] S5, selecting a supplementary point in the triangulated area of ​​the image plane, calculating the X-axis coordinate and the Y-axis coordinate of the supplementary point in the point cloud coordinate system by using the collinear equation, substituting them into the fitting plane equation, calculating the elevation of the supplementary point, and adding the corresponding supplementary point to the point cloud data according to the X-axis coordinate, the Y-axis coordinate and the elevation;

[0028] Specifically, the supplementary point can be a point, such as the centroid of the triangular area, or multiple points, such as a series of points equidistant from the centroid of the triangular area. Taking the centroid as an example, the image plane coordinates of the centroid are known. Substituting them into the collinear equation, we can get its X-axis coordinate and Y-axis coordinate in the point cloud coordinate system. Substituting the X-axis coordinate and the Y-axis coordinate into the plane equation, we can get the Z-axis coordinate, that is, the elevation, and thus add the point to the point cloud data.

[0029] S6, repeating S3 to S5, adding supplementary points to multiple sub-triangulated networks of the point cloud data;

[0030] Specifically, additional points can be added to each sub-triangle network or part of the sub-triangle network to expand the points in the point cloud and increase the density of points in the point cloud data. Therefore, S3 to S6 are used to increase the point density, which saves costs compared to the existing method of increasing the point density during acquisition to increase the detail texture.

[0031] S7. Assign true color information to the added supplementary points based on the RGB information of the supplementary points in the image plane.

[0032] Specifically, true color information is assigned to the supplementary points according to the existing coloring technology of points in point cloud data, so that detailed textures can be clearly observed in the point cloud data visualization interface.

Claims

1. A method for supplementing detail texture for point cloud using images, characterized in that: The following steps are involved: S1, obtain point cloud data and its corresponding images; S2, establish a Delaunay triangulation network based on point cloud data; S3, selecting a sub-triangular network, fitting a fitting plane of the sub-triangular network, and obtaining a fitting plane equation; S4, projecting the three vertices of the sub-triangle network onto the image plane respectively through collinear equations to obtain the image plane triangular region corresponding to the sub-triangle network; S5, selecting a supplementary point in the triangulated area of ​​the image plane, calculating the X-axis coordinate and the Y-axis coordinate of the supplementary point in the point cloud coordinate system by using the collinear equation, substituting them into the fitting plane equation, calculating the elevation of the supplementary point, and adding the corresponding supplementary point to the point cloud data according to the X-axis coordinate, the Y-axis coordinate and the elevation; S6, repeating S3 to S5, adding supplementary points to multiple sub-triangulated networks of the point cloud data; S7. Assign true color information to the added supplementary points based on the RGB information of the supplementary points in the image plane.

2. The method for supplementing detail texture for point cloud using image according to claim 1, characterized in that: The fitting plane equation is Z=AX+BY+C, where A, B and C are parameters, which are obtained by substituting the coordinate values ​​of the three vertices into the equation, or solved by the least squares method, Lagrange multiplier method or SVD decomposition method.

3. The method for supplementing detail texture for point cloud using image according to claim 1, characterized in that: The collinear equation is: ,in,( , ) represents the coordinates of the vertex in the image plane, is the focal length of the camera, (X, Y, Z) represents the coordinates of the vertex in the point cloud coordinate system, ( , , ) represents the coordinates of the camera taking the image in the point cloud coordinate system, , , , , , , , and It is the direction cosines determined by the three angular elements of the image.

4. The method for supplementing detail texture for point cloud using image according to any one of claims 1 to 3, characterized in that: In S5, the supplementary point includes the centroid of the triangular region or a series of points equidistant from the centroid of the triangular region.