Point cloud completion method and system based on secondary Poisson reconstruction

Abstract

The invention relates to the technical field of point cloud completion, in particular to a point cloud completion method and system based on secondary Poisson reconstruction, and the method comprises the steps: obtaining three-dimensional space coordinates and color information obtained through the analysis and recognition of original high-density point cloud data, converting the three-dimensional space coordinates and color information into a uniform data format, and carrying out the preprocessing; poisson surface reconstruction is carried out on the to-be-completed point cloud data based on a Poisson reconstruction algorithm and an Open3D framework, wherein the Poisson surface reconstruction comprises normal vector estimation, density-driven grid optimization and implicit surface generation; performing secondary reconstruction on the primary Poisson reconstruction point cloud data based on a secondary Poisson reconstruction algorithm; and acquiring small-range point cloud data around the missing part of the point cloud profile obtained by cutting based on FAROSCENE software, and comparing and fitting the original high-density point cloud data, the small-range point cloud data around the missing part and the secondary Poisson reconstruction point cloud data. The method can greatly improve the complementation precision, and has high fidelity and high visualization quality.

Description

Technical Field

[0001] The present invention relates to the technical field of point cloud completion, and in particular to a point cloud completion method and system based on quadratic Poisson reconstruction. Background Art

[0002] The preservation of geological outcrops faces the risk of irreversible damage. Traditional manual surveying methods suffer from low precision (millimeter-level errors), poor efficiency (time-consuming single-point measurements), and poor integrity (two-dimensional records cannot represent three-dimensional structures), making them inadequate for modern geological research. Therefore, 3D laser scanners have been introduced for field geological modeling and preservation. While 3D laser scanning technology can achieve submillimeter-level accuracy (typical accuracy of 0.3mm for FAROFocus equipment) in high-density field point cloud modeling, it is limited by complex geological structural features (such as bedding dislocations, faults, and fissures) and environmental occlusion (such as fences and vegetation cover). Point cloud data generally suffers from partial missing data of 15%-30%, severely impacting the integrity of geological model reconstruction. Therefore, point cloud completion is necessary to address these missing areas. The improved point cloud completion algorithm based on Poisson reconstruction uses secondary Poisson reconstruction and improved point cloud fusion. It not only solves the unevenness caused by the traditional single geometric algorithm's inability to handle the variable normal vectors and complex curvature in geological outcrops, but also optimizes the topological relationship errors in completing large-scale missing areas, greatly improving the three-dimensional modeling effect and digital preservation capabilities of precious geological outcrops in the field.

[0003] However, the existing technology has the following shortcomings: 1. Using only traditional single geometric algorithms cannot handle the variable normal vector directions and complex curvatures in geological outcrops, resulting in serious non-smooth protrusions in the completed profiles; 2. Existing technologies are prone to topological errors when reconstructing the mesh when filling in large missing areas, resulting in extremely poor completion results. 3. Mainstream deep learning methods require at least 10^4 training samples, but 3D point cloud data of geological outcrops is difficult to obtain, and direct application of less data will lead to overfitting (excessive error in the validation set). Summary of the Invention

[0004] The purpose of the present invention is to provide a point cloud completion method and system based on quadratic Poisson reconstruction to solve the problems in the prior art of serious non-smooth protrusion of the completed profile, easy topological errors and excessive completion errors.

[0005] To achieve the above object, the present invention provides a point cloud completion method based on quadratic Poisson reconstruction, comprising the following steps: S1. Obtaining the three-dimensional spatial coordinates and color information obtained by analyzing and identifying the original high-density point cloud data, converting the three-dimensional spatial coordinates and color information into a unified data format and performing preprocessing to obtain the point cloud data to be completed; S2. Based on the Poisson reconstruction algorithm and the Open3D framework, Poisson surface reconstruction is performed on the point cloud data to be completed, including normal vector estimation, density-driven mesh optimization and implicit surface generation, to obtain a Poisson reconstructed point cloud data; S3, performing secondary reconstruction on the primary Poisson reconstruction point cloud data based on a secondary Poisson reconstruction algorithm to obtain secondary Poisson reconstruction point cloud data; S4. Obtain the small-scale point cloud data around the missing part of the point cloud section obtained by cropping with FAROSCENE software, compare and fit the original high-density point cloud data, the small-scale point cloud data around the missing part, and the quadratic Poisson reconstruction point cloud data to obtain the point cloud completion result.

[0006] In some embodiments of the present application, in S1, the three-dimensional space coordinates and color information are converted into a unified data format and pre-processed to obtain the point cloud data to be completed, including: Convert 3D spatial coordinates and color information into a unified data format, read them using LAS and PLY file-compatible reading logic, and automatically identify file extensions using the os.path.splitext method to distinguish data formats. Also, check the integrity of the data content. The three-dimensional spatial coordinates and color information in a unified data format are formatted, repaired, and normalized in sequence to obtain the point cloud data to be completed.

[0007] In some embodiments of the present application, in S2, performing Poisson surface reconstruction on the to-be-completed point cloud data based on the Poisson reconstruction algorithm and the Open3D framework includes: In the Open3D framework, a hybrid strategy of radius search and nearest neighbor number constraint is adopted, and a forced normal vector with a modulus of 1 is called; The density value densities output by Poisson reconstruction reflects the vertex confidence and follows a long-tail distribution. The 5% quantile threshold is calculated to dynamically identify low-density vertices. Based on the density mask, low-density vertices are removed and high-confidence areas are retained, so that the mesh topology conforms to the preset geological laws. The generated mesh model is evenly sampled to obtain a Poisson reconstructed point cloud data.

[0008] In some embodiments of the present application, in S3, performing secondary reconstruction on the primary Poisson reconstruction point cloud data based on the secondary Poisson reconstruction algorithm to obtain the secondary Poisson reconstruction point cloud data includes: Based on the secondary Poisson reconstruction algorithm, the protruding part of the primary Poisson reconstruction point cloud data is downsampled again to reduce the point cloud density of the protruding part and the redundant points formed by the outward expansion of the edge points of the point cloud in the primary Poisson reconstruction point cloud data, thereby obtaining the secondary Poisson reconstruction point cloud data.

[0009] In some embodiments of the present application, in S4, the original high-density point cloud data, the small-area point cloud data around the missing part, and the quadratic Poisson reconstructed point cloud data are compared and fitted to obtain a point cloud completion result including: The original high-density point cloud data, the small-area point cloud data around the missing part, and the high-density part of the quadratic Poisson reconstructed point cloud data are fitted, and the low-density part is eliminated to obtain the point cloud completion result.

[0010] In some embodiments of the present application, S4 further includes: ContextCapture software is used to visualize the three-dimensional model of the point cloud completion results.

[0011] In some embodiments of the present application, a point cloud completion system based on quadratic Poisson reconstruction is also disclosed, including: The data acquisition module is used to obtain the three-dimensional spatial coordinates and color information obtained by analyzing and identifying the original high-density point cloud data, convert the three-dimensional spatial coordinates and color information into a unified data format, and perform preprocessing to obtain the point cloud data to be completed; The single-stage reconstruction module is used to perform Poisson surface reconstruction on the completed point cloud data based on the Poisson reconstruction algorithm and the Open3D framework, including normal vector estimation, density-driven mesh optimization, and implicit surface generation, to obtain the single-stage Poisson reconstructed point cloud data; A secondary reconstruction module is used to perform secondary reconstruction on the primary Poisson reconstruction point cloud data based on a secondary Poisson reconstruction algorithm to obtain secondary Poisson reconstruction point cloud data; The comparison and fitting module is used to obtain the small-scale point cloud data around the missing part of the point cloud section cropped by FAROSCENE software, and compare and fit the original high-density point cloud data, the small-scale point cloud data around the missing part, and the quadratic Poisson reconstruction point cloud data to obtain the point cloud completion result.

[0012] The advantages and beneficial effects of the present invention over the prior art are: 1. This application uses secondary Poisson reconstruction to further optimize the geometric structure and topological relationships of the point cloud. Compared to primary Poisson reconstruction, secondary reconstruction can better smooth the point cloud surface and reduce geometric mutations caused by noise and irregular sampling, thereby improving the overall quality of the completed point cloud. 2. The application's quadratic Poisson reconstruction, through sophisticated mathematical models and algorithm optimization during implicit surface generation and mesh optimization, can effectively smooth point cloud surfaces and better handle noise and irregular sampling points in point cloud data, avoiding geometric mutations and non-smoothness caused by these factors. 3. During the comparison and fitting stage, this application conducts a comprehensive analysis and fitting of the original high-density point cloud data, the small-scale point cloud data around the missing part, and the secondary Poisson reconstruction point cloud data, which can more accurately locate the completion area and ensure the geometric and topological consistency of the completed point cloud with the original data, thereby improving the accuracy of the completion.

[0013] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 Schematic diagram of the steps of a point cloud completion method based on quadratic Poisson reconstruction in an embodiment of the present invention; Figure 2 This is a general flow chart of the code for the improved point cloud completion algorithm according to an embodiment of the present invention; Figure 3 2 is a structural diagram of a point cloud completion system based on quadratic Poisson reconstruction in an embodiment of the present invention; Figure 4 A point cloud model diagram of a geological outcrop partially obscured by a railing according to an embodiment of the present invention; Figure 5 This is a point cloud model diagram of a geological outcrop after removing the railings according to an embodiment of the present invention; Figure 6 A small-area point cloud data diagram around the missing part of an embodiment of the present invention; Figure 7 This is a diagram showing the point cloud completion result according to an embodiment of the present invention; Figure 8 This is a Mesh model rendering of the point cloud completion result of an embodiment of the present invention. DETAILED DESCRIPTION

[0015] In the description of the present invention, it should be noted that the terms "upper", "lower", "inside", "outside" and the like indicate orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, or the orientations or positional relationships in which the inventive product is usually placed when in use. These are only for the convenience of describing the present invention and simplifying the description, and are not intended to indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation. Therefore, they cannot be understood as limitations on the present invention. In the description of the present invention, it should also be noted that, unless otherwise expressly specified and limited, the terms "setting", "installation" and "connection" should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or an indirect connection through an intermediate medium, or it can be a communication between the internal parts of two elements. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0016] The embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0017] like Figure 1 As shown, the present invention provides a point cloud completion method based on quadratic Poisson reconstruction, comprising the following steps: S1. Obtaining the three-dimensional spatial coordinates and color information obtained by analyzing and identifying the original high-density point cloud data, converting the three-dimensional spatial coordinates and color information into a unified data format and performing preprocessing to obtain the point cloud data to be completed; S2. Based on the Poisson reconstruction algorithm and the Open3D framework, Poisson surface reconstruction is performed on the point cloud data to be completed, including normal vector estimation, density-driven mesh optimization and implicit surface generation, to obtain a Poisson reconstructed point cloud data; S3, performing secondary reconstruction on the primary Poisson reconstruction point cloud data based on a secondary Poisson reconstruction algorithm to obtain secondary Poisson reconstruction point cloud data; S4. Obtain the small-scale point cloud data around the missing part of the point cloud section obtained by cropping with FAROSCENE software, compare and fit the original high-density point cloud data, the small-scale point cloud data around the missing part, and the quadratic Poisson reconstruction point cloud data to obtain the point cloud completion result.

[0018] It should be understood that Poisson reconstruction uses implicit surface fitting technology to restore continuous and smooth geological structure surfaces from residual point clouds, but a single Poisson reconstruction algorithm has problems with over-smoothing of surfaces and loss of details in the completion of geological outcrop point clouds, making it difficult to maintain geometric sharpness. This application proposes an improved Poisson reconstruction algorithm, which uses a point cloud merging method with secondary Poisson reconstruction and local geometric feature enhancement, integrates high-density laser scanning data with uniformly sampled data after secondary reconstruction, and combines the small-scale point cloud data around the missing part. The three are compared and fused, the high-density part is fitted, and the low-density point cloud of the same part is eliminated to achieve geometric and color consistency reconstruction of the missing area. Significantly improve the reconstruction accuracy of complex geological structures. The algorithm of this application is implemented through the following process code, and the overall code process is as follows: Figure 2 shown.

[0019] In some embodiments of the present application, in S1, the three-dimensional space coordinates and color information are converted into a unified data format and pre-processed to obtain the point cloud data to be completed, including: Convert 3D spatial coordinates and color information into a unified data format, read them using LAS and PLY file-compatible reading logic, and automatically identify file extensions using the os.path.splitext method to distinguish data formats. Also, check the integrity of the data content. The three-dimensional spatial coordinates and color information in a unified data format are formatted, repaired, and normalized in sequence to obtain the point cloud data to be completed.

[0020] Specifically, reading point cloud data involves parsing the original file, extracting 3D coordinates and color information, and converting it into a unified data representation. To support different data formats for point cloud reconstruction, the code includes LAS and PLY file reading logic. The os.path.splitext method automatically identifies file extensions to distinguish data formats, and checks the integrity of file contents. For example, in the case of PLY point clouds, the code ensures file validity. If the file is unavailable or empty, the logic will throw a clear error message to avoid subsequent calculation problems.

[0021] In some embodiments of the present application, in S2, performing Poisson surface reconstruction on the to-be-completed point cloud data based on the Poisson reconstruction algorithm and the Open3D framework includes: In the Open3D framework, a hybrid strategy of radius search and nearest neighbor number constraint is adopted, and a forced normal vector with a modulus of 1 is called; The density value densities output by Poisson reconstruction reflects the vertex confidence and follows a long-tail distribution. The 5% quantile threshold is calculated to dynamically identify low-density vertices. Based on the density mask, low-density vertices are removed and high-confidence areas are retained, so that the mesh topology conforms to the preset geological laws. The generated mesh model is evenly sampled to obtain a Poisson reconstructed point cloud data.

[0022] Specifically, this solution uses a hybrid strategy combining a radius search (radius=0.2) and a nearest neighbor constraint (max_nn=30). Normalize_normals() is called to enforce a modulus of 1 for the normal vectors, preventing weight bias in the subsequent Poisson equation solution. The density values ​​(densities) output by the Poisson reconstruction reflect vertex confidence and follow a long-tailed distribution. Low-probability vertices (noise or outliers) are dynamically identified by calculating a 5% quantile threshold (np.percentile(densities,5)). Low-density vertices (remove_vertices_by_mask) are removed based on a density mask (vertices_to_remove), retaining high-confidence areas to ensure that the mesh topology conforms to geological laws. The resulting mesh is then uniformly sampled.

[0023] In some embodiments of the present application, in S3, performing secondary reconstruction on the primary Poisson reconstruction point cloud data based on the secondary Poisson reconstruction algorithm to obtain the secondary Poisson reconstruction point cloud data includes: Based on the secondary Poisson reconstruction algorithm, the protruding part of the primary Poisson reconstruction point cloud data is downsampled again to reduce the point cloud density of the protruding part and the redundant points formed by the outward expansion of the edge points of the point cloud in the primary Poisson reconstruction point cloud data, thereby obtaining the secondary Poisson reconstruction point cloud data.

[0024] It is important to understand that secondary Poisson reconstruction is an iterative optimization algorithm proposed based on primary Poisson reconstruction, which aims to solve the problem of detail loss caused by noise interference, normal vector deviation or complex geometric features in the first reconstruction. The core of this solution is to downsample the protrusions formed after the first reconstruction through secondary Poisson reconstruction, reduce the point cloud density of the protrusions, and thus improve the reconstruction accuracy of complex geological structures. The uniform sampling after the secondary Poisson reconstruction can effectively reduce the redundant points formed by the expansion of the edge points of the completed point cloud during the first Poisson reconstruction, reduce the protrusions in the completed periphery, and improve the smoothness and fluidity of the entire profile.

[0025] In some embodiments of the present application, in S4, the original high-density point cloud data, the small-area point cloud data around the missing part, and the quadratic Poisson reconstructed point cloud data are compared and fitted to obtain a point cloud completion result including: The original high-density point cloud data, the small-area point cloud data around the missing part, and the high-density part of the quadratic Poisson reconstructed point cloud data are fitted, and the low-density part is eliminated to obtain the point cloud completion result.

[0026] It should be understood that this solution introduces a small range of point cloud data around the missing part, which has obvious advantages in point cloud data fusion. Because Poisson reconstruction is a point cloud completion based on the original data of the entire section, the use of small-scale point cloud data merging can optimize the Poisson reconstruction algorithm to complete the defect of its own boundary expansion; because the original point cloud file is large, point cloud block completion is used in the specific operation process. When traversing to check whether the missing points need to be inserted, the small-scale point cloud improves the traversal efficiency and saves computer computing time and storage space; in the process of inserting point color assignment, the point selection on the small-scale point cloud data is not affected by the color of the entire section, making the completed point cloud color more accurate.

[0027] In some embodiments of the present application, S4 further includes: ContextCapture software is used to visualize the three-dimensional model of the point cloud completion results.

[0028] In some embodiments of the present application, Figure 3 As shown, a point cloud completion system based on quadratic Poisson reconstruction is also disclosed, including: The data acquisition module is used to obtain the three-dimensional spatial coordinates and color information obtained by analyzing and identifying the original high-density point cloud data, convert the three-dimensional spatial coordinates and color information into a unified data format, and perform preprocessing to obtain the point cloud data to be completed; The single-stage reconstruction module is used to perform Poisson surface reconstruction on the completed point cloud data based on the Poisson reconstruction algorithm and the Open3D framework, including normal vector estimation, density-driven mesh optimization, and implicit surface generation, to obtain the single-stage Poisson reconstructed point cloud data; A secondary reconstruction module is used to perform secondary reconstruction on the primary Poisson reconstruction point cloud data based on a secondary Poisson reconstruction algorithm to obtain secondary Poisson reconstruction point cloud data; The comparison and fitting module is used to obtain the small-scale point cloud data around the missing part of the point cloud section cropped by FAROSCENE software, and compare and fit the original high-density point cloud data, the small-scale point cloud data around the missing part, and the quadratic Poisson reconstruction point cloud data to obtain the point cloud completion result.

[0029] The advantages and beneficial effects of the present invention over the prior art are: 1. This application uses secondary Poisson reconstruction to further optimize the geometric structure and topological relationships of the point cloud. Compared to primary Poisson reconstruction, secondary reconstruction can better smooth the point cloud surface and reduce geometric mutations caused by noise and irregular sampling, thereby improving the overall quality of the completed point cloud. 2. The application's quadratic Poisson reconstruction, through sophisticated mathematical models and algorithm optimization during implicit surface generation and mesh optimization, can effectively smooth point cloud surfaces and better handle noise and irregular sampling points in point cloud data, avoiding geometric mutations and non-smoothness caused by these factors. 3. During the comparison and fitting stage, this application conducts a comprehensive analysis and fitting of the original high-density point cloud data, the small-scale point cloud data around the missing part, and the secondary Poisson reconstruction point cloud data, which can more accurately locate the completion area and ensure the geometric and topological consistency of the completed point cloud with the original data, thereby improving the accuracy of the completion.

[0030] The following describes the implementation of the present invention in detail with reference to specific examples.

[0031] In one embodiment, one of the geological outcrops in the Longhuitou Scenic Area is blocked by a publicity fence, such as Figure 4 As shown in the figure, it cannot fully display all the information of the geological outcrop. The railings of the geological outcrop were removed by the cropping function of the FAROSCENE software. After the removal, a large part of the point cloud data was left missing. The geological outcrop after removing the railings is as follows Figure 5 shown.

[0032] After the second Poisson reconstruction, when merging the point clouds, the introduction of the section missing small range point cloud data such as Figure 6 As shown, the original high-density point cloud data and the uniformly sampled data after secondary reconstruction are compared and fused to generate a complete geological outcrop profile, as shown in Figure 7 As shown, the 3D model of the completion result is visualized by ContextCapture software, as shown Figure 8 shown.

[0033] In summary, this improved point cloud completion algorithm based on quadratic Poisson reconstruction achieves high point cloud density, smooth surfaces, and high visualization quality in areas where point clouds are missing. While preserving key geological features, the point cloud achieves a balance between the number of points (4,239,124) and the number of triangles (39,564), validating its advantages in high-precision reconstruction. It performs exceptionally well in geological structure reconstruction, achieving high completion accuracy, good mesh modeling, and smooth surfaces. Although some protrusions are present, this significantly improves model accuracy. Its high fidelity and visualization quality provide reliable technical support for the scientific protection and cultural heritage of precious geological resources.

[0034] In this application, unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which this application belongs. In the event of any inconsistency, the meaning described in this specification or the meaning derived from the contents recorded in this specification shall prevail. In addition, the terms used herein are only for the purpose of describing the embodiments of this application and are not intended to limit this application.

[0035] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit the same. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that they can still modify or replace the technical solutions of the present invention with equivalents, and these modifications or equivalent replacements cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A point cloud completion method based on quadratic Poisson reconstruction, characterized in that: The following steps are involved: S1. Obtaining the three-dimensional spatial coordinates and color information obtained by analyzing and identifying the original high-density point cloud data, converting the three-dimensional spatial coordinates and color information into a unified data format and performing preprocessing to obtain the point cloud data to be completed; S2. Based on the Poisson reconstruction algorithm and the Open3D framework, Poisson surface reconstruction is performed on the point cloud data to be completed, including normal vector estimation, density-driven mesh optimization and implicit surface generation, to obtain a Poisson reconstructed point cloud data; S3, performing secondary reconstruction on the primary Poisson reconstruction point cloud data based on a secondary Poisson reconstruction algorithm to obtain secondary Poisson reconstruction point cloud data; S4. Obtain the small-scale point cloud data around the missing part of the point cloud section obtained by cropping with FAROSCENE software, compare and fit the original high-density point cloud data, the small-scale point cloud data around the missing part, and the quadratic Poisson reconstruction point cloud data to obtain the point cloud completion result.

2. The point cloud completion method based on quadratic Poisson reconstruction according to claim 1, characterized in that: In S1, the three-dimensional space coordinates and color information are converted into a unified data format and pre-processed to obtain the point cloud data to be completed, including: Convert 3D spatial coordinates and color information into a unified data format, read them using LAS and PLY file-compatible reading logic, and automatically identify file extensions using the os.path.splitext method to distinguish data formats. Also, check the integrity of the data content. The three-dimensional spatial coordinates and color information in a unified data format are formatted, repaired, and normalized in sequence to obtain the point cloud data to be completed.

3. The point cloud completion method based on quadratic Poisson reconstruction according to claim 2, characterized in that: In S2, the Poisson surface reconstruction of the completed point cloud data based on the Poisson reconstruction algorithm and the Open3D framework includes: In the Open3D framework, a hybrid strategy of radius search and nearest neighbor number constraint is adopted, and a forced normal vector with a modulus of 1 is called; The density value densities output by Poisson reconstruction reflects the vertex confidence and follows a long-tail distribution. The 5% quantile threshold is calculated to dynamically identify low-density vertices. Based on the density mask, low-density vertices are removed and high-confidence areas are retained, so that the mesh topology conforms to the preset geological laws. The generated mesh model is evenly sampled to obtain a Poisson reconstructed point cloud data.

4. The point cloud completion method based on quadratic Poisson reconstruction according to claim 3, characterized in that: In S3, the primary Poisson reconstruction point cloud data is reconstructed for the secondary time based on the secondary Poisson reconstruction algorithm, and the secondary Poisson reconstruction point cloud data obtained includes: Based on the secondary Poisson reconstruction algorithm, the protruding part of the primary Poisson reconstruction point cloud data is downsampled again to reduce the point cloud density of the protruding part and the redundant points formed by the outward expansion of the edge points of the point cloud in the primary Poisson reconstruction point cloud data, thereby obtaining the secondary Poisson reconstruction point cloud data.

5. The point cloud completion method based on quadratic Poisson reconstruction according to claim 4, characterized in that: In S4, the original high-density point cloud data, the small-area point cloud data around the missing part, and the quadratic Poisson reconstruction point cloud data are compared and fitted to obtain the point cloud completion results including: The original high-density point cloud data, the small-area point cloud data around the missing part, and the high-density part of the quadratic Poisson reconstructed point cloud data are fitted, and the low-density part is eliminated to obtain the point cloud completion result.

6. The point cloud completion method based on quadratic Poisson reconstruction according to claim 5, characterized in that: Said S4 further includes: ContextCapture software is used to visualize the three-dimensional model of the point cloud completion results.

7. A point cloud completion system based on quadratic Poisson reconstruction, comprising: The data acquisition module is used to obtain the three-dimensional spatial coordinates and color information obtained by analyzing and identifying the original high-density point cloud data, convert the three-dimensional spatial coordinates and color information into a unified data format, and perform preprocessing to obtain the point cloud data to be completed; The single-stage reconstruction module is used to perform Poisson surface reconstruction on the completed point cloud data based on the Poisson reconstruction algorithm and the Open3D framework, including normal vector estimation, density-driven mesh optimization, and implicit surface generation, to obtain the single-stage Poisson reconstructed point cloud data; A secondary reconstruction module is used to perform secondary reconstruction on the primary Poisson reconstruction point cloud data based on a secondary Poisson reconstruction algorithm to obtain secondary Poisson reconstruction point cloud data; The comparison and fitting module is used to obtain the small-scale point cloud data around the missing part of the point cloud section cropped by FAROSCENE software, and compare and fit the original high-density point cloud data, the small-scale point cloud data around the missing part, and the quadratic Poisson reconstruction point cloud data to obtain the point cloud completion result.

Images