Disordered three-dimensional discrete point cloud slice type feature extraction method and system

Abstract

The invention discloses an unordered three-dimensional discrete point cloud slice type feature extraction method and system. The method comprises the steps that the effective range and density of disordered three-dimensional discrete points are acquired, and intervals of slices are set in the X direction, the Y direction and the Z direction according to the density; the disordered three-dimensional discrete points are sliced from small to large in the X direction, the Y direction and the Z direction according to the set intervals of the slices, and two-dimensional slices in all the directions are obtained; according to the two-dimensional slice data in each direction, respectively calculating a concave packet of each tangent plane in the X, Y and Z directions, and marking points contained in the outer boundary of the concave packet; counting the repeatedly marked points to form a repeatedly marked point set; and carrying out voxel filtering on the repeatedly marked point set to obtain extracted feature points. The method is inspired by medical CT slices, three-dimensional entities of any shape and size can be processed, and the problems that an existing disordered three-dimensional discrete point cloud feature extraction algorithm is complex and time-consuming in processing can be solved.

Description

Technical Field

[0001] The present invention belongs to the field of laser point cloud feature extraction, and in particular relates to a method for extracting feature points of a three-dimensional disordered discrete point cloud by continuously slicing and segmenting the three-dimensional discrete point cloud in the x, y, and z directions to obtain a concave hull and mark it, thereby realizing feature point extraction of a three-dimensional disordered discrete point cloud. Specifically, it is a slicing feature extraction method and system for disordered three-dimensional discrete point clouds. Background Art

[0002] With the rapid development of 3D acquisition technology, the acquisition of disordered 3D discrete point cloud data is becoming increasingly convenient. Disordered 3D discrete point cloud is the basic data source for 3D spatial information processing, but due to factors such as acquisition environment, method and sensor positioning, data collected multiple times often cannot be directly spliced. Therefore, point cloud registration (PCR) technology came into being and became the key to achieving accurate data alignment. In the point cloud data processing task, feature extraction, as an important pre-processing step, provides important support for subsequent feature matching, registration and overall alignment, and the two are closely related. Efficient and accurate feature extraction is the basis of registration quality. The existing point feature extraction methods are mainly SIFT3D, Harris3D, NARF, etc. Although these methods can extract point features very well. However, they all have a common disadvantage, that is, the algorithms are relatively complex, the processing is time-consuming, and there are requirements for point clouds, which limits the rapid processing of 3D geographic information data. Summary of the invention

[0003] In order to overcome the shortcomings of the above-mentioned prior art, the present invention provides a slicing feature extraction method and system for disordered three-dimensional discrete point clouds, which is inspired by medical CT slices and can process three-dimensional entities of any shape and size, and can solve the problems of complex and time-consuming processing of existing disordered three-dimensional discrete point cloud feature extraction algorithms.

[0004] According to one aspect of the present invention, a method for extracting features by slicing an unordered three-dimensional discrete point cloud is provided, comprising: Get the effective range and density of disordered 3D discrete points, and set the interval of slices in X, Y, and Z directions according to the density; According to the set slice interval, the disordered three-dimensional discrete points are sliced ​​from small to large in the three directions of X, Y, and Z to obtain two-dimensional slices in each direction; According to the two-dimensional slice data in each direction, the concave hull of each slice in the X, Y, and Z directions is calculated respectively, and the points contained in the outer boundary of the concave hull are marked; Count the repeatedly marked points to form a repeatedly marked point set; Voxel filtering is performed on the repeatedly marked point set to obtain the extracted feature points.

[0005] As a further technical solution, the disordered three-dimensional discrete points are sliced ​​from small to large in the three directions of X, Y, and Z, including: Count the maximum and minimum values ​​of disordered three-dimensional discrete points in the X, Y, and Z directions; Use the intervals dx, dy, and dz to slice the disordered three-dimensional discrete points from small to large in the corresponding directions of X, Y, and Z.

[0006] As a further technical solution, two-dimensional slices in various directions are obtained, including: A plane coordinate system is established for the section with two coordinate axes perpendicular to the slicing direction, and the slicing direction coordinate is set to 0 to reduce the 3D point cloud to a 2D point cloud.

[0007] As a further technical solution, the concave hulls of each section in the X, Y, and Z directions are calculated respectively, and the points contained in the outer boundary of the concave hull are marked, including: Use the classic rolling ball method to find the concave hull of each section; For the points on the concave hull obtained, the subscripts in the original unordered three-dimensional discrete point cloud are found through the KD tree and marked.

[0008] As a further technical solution, the concave hull of each section is obtained using the classic rolling ball method, including: Find the polar center P0 of the two-dimensional point set; Find the leftmost point as the first point P1 of the concave edge; The radius of the ball set by the rolling ball method is used to filter out the point set Pnearest within the fixed radius with P1 as the center; Traverse Pnearest, take the angle P0->P1 as the reference, rotate counterclockwise to find the point with the smallest counterclockwise angle and add it to the concave hull; Repeat the above steps until no new points are found or a point that has been previously added to the concave hull is encountered.

[0009] As a further technical solution, voxel filtering is performed on the repeatedly marked point set, including: The repeatedly marked point set is voxelized with voxel radius to obtain a set of feature points with consistent resolution and uniform distribution.

[0010] According to one aspect of the present invention, a system for extracting features from an unordered three-dimensional discrete point cloud slice is provided, comprising: The first main module is used to obtain the effective range and density of disordered three-dimensional discrete points, and set the interval of slices in the X, Y and Z directions according to the density; The second main module is used to slice the disordered three-dimensional discrete points from small to large in the three directions of X, Y and Z according to the set slice interval to obtain two-dimensional slices in each direction; The third main module is used to calculate the concave hull of each section in the X, Y and Z directions according to the two-dimensional slice data in each direction, and mark the points contained in the outer boundary of the concave hull; The fourth main module is used to count the repeatedly marked points to form a repeatedly marked point set; The fifth main module is used to perform voxel filtering on the repeatedly marked point set to obtain extracted feature points.

[0011] According to one aspect of the present invention specification, there is provided an electronic device, comprising a processor and a memory; the memory stores program instructions executed by the processor, and the processor calls the program instructions to execute the steps of the method for slicing feature extraction of an unordered three-dimensional discrete point cloud.

[0012] According to one aspect of the present specification, a non-transitory computer-readable storage medium is provided, wherein the non-transitory computer-readable storage medium stores computer instructions, wherein the computer instructions enable the computer to execute the steps of the method for extracting features from an unordered three-dimensional discrete point cloud by slicing.

[0013] Compared with the prior art, the present invention has the following beneficial effects: (1) The present invention is inspired by CT slices in medicine. The method slices the point cloud, then marks possible feature points according to the edge of the smallest concave polygon on the slice, confirms the feature points by repeated marking, and finally avoids dense feature points by voxel filtering. Compared with some classic feature point extraction algorithms such as SIFT, Harris3D, and NARF, the method of the present invention is simpler and more direct in algorithm implementation. It does not need to calculate the gradient and normal direction of the point cloud data, but only determines the feature points by counting the edge points of the concave hull of each slice, and has a high processing efficiency.

[0014] (2) The advantage of the method of the present invention is that the calculation is simple and does not rely on more complex geometric properties such as calculating derivatives. Therefore, it is particularly suitable for embedded devices with limited computing power but need to process point cloud features in real time and efficiently. For example, low-power microcontrollers, unmanned vehicles, robots or DSP systems in the field of industrial control. Such lightweight devices have higher requirements for algorithm complexity in point cloud preprocessing. This method realizes feature point recognition through extremely simple calculations, maximizing processing speed while ensuring the necessary feature quality, which is why it has a strong advantage in application promotion. It fills the technical gap in real-time low-power device feature extraction. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, a brief introduction is given below to the drawings used in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0016] Figure 1 A schematic flow chart of a method for extracting features from a slicing-based unordered three-dimensional discrete point cloud provided in an embodiment of the present invention.

[0017] Figure 2 (a)-(c) are schematic diagrams of the slicing principle provided in an embodiment of the present invention.

[0018] Figure 3 A schematic diagram of obtaining a concave hull using the rolling ball method provided in an embodiment of the present invention.

[0019] Figure 4 A schematic diagram of an example of feature extraction results provided by an embodiment of the present invention.

[0020] Figure 5 A structural schematic diagram of an unordered three-dimensional discrete point cloud slicing feature extraction system provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0021] In view of the current situation that the existing algorithms are relatively complex, time-consuming to process, and have requirements for point clouds, which limits the rapid processing of three-dimensional geographic information data, the present invention is inspired by CT slices in medicine and proposes a slicing feature extraction algorithm for disordered three-dimensional discrete point clouds. CT slices in medicine can decompose any entity into slices. The algorithm proposed in the present invention is similar to the slicing process. First, the slicing operation and dimensionality reduction are carried out on the three directions of x, y, and z of the disordered three-dimensional discrete points. Then, the minimum concave polygon is found on the slice and the edge points are marked. Finally, the point set marked multiple times is voxel filtered to obtain the final feature points. The algorithm has a simple processing process, high computational efficiency, no requirements for data points, and can process three-dimensional entities of any shape and size. It is particularly suitable for rapid feature extraction of three-dimensional geographic information data. This method avoids complex three-dimensional spatial feature calculations. The calculations in each slice are simple, and the large number of independent subtasks formed by slicing are also more conducive to parallel processing, which can stably describe point clouds.

[0022] The terms "including" and "having" and any variations thereof in the specification and claims of the present invention and the above-mentioned drawings are intended to cover non-exclusive inclusions, for example, a process, method, system, product or apparatus comprising a series of steps or units is not necessarily limited to the steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to these processes, methods, products or apparatuses.

[0023] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention. In addition, the technical features in the various embodiments or single embodiments provided by the present invention are arbitrarily combined with each other to form a new technical solution. This combination is not restricted by the sequence of steps and / or the structural composition mode, but must be based on the ability of ordinary technicians in this field to achieve. When the combination of technical solutions is contradictory or cannot be achieved, it should be considered that this combination of technical solutions does not exist and is not within the scope of protection required by the present invention.

[0024] Inspired by CT slices in daily life, the present invention proposes a slicing feature extraction method for disordered three-dimensional discrete point clouds. The core of the method includes three processes: first, slicing operations are performed on disordered three-dimensional discrete points, then the smallest concave polygon is found on the section and the edge points are marked, and finally, the point set marked multiple times is subjected to voxel filtering to obtain the final feature points. The method has a simple processing process, avoids complex three-dimensional spatial feature calculations, has high computational efficiency, has no requirements for data points, can process three-dimensional entities of any shape and size, and a large number of independent subtasks formed by slicing are also more conducive to parallel processing, which is particularly suitable for fast feature extraction of three-dimensional geographic information data.

[0025] like Figure 1 As shown, a slicing feature extraction method of an unordered three-dimensional discrete point cloud according to an embodiment of the present invention comprises the following steps: Step 1: Get the valid range of disordered three-dimensional discrete points and X,Y,Z Set appropriate density values ​​in three directions dx, dy, dz is the interval, in the corresponding direction (such as X ), slice the disordered three-dimensional discrete points from small to large. The schematic diagram of the slicing principle is shown in Figure 2 As shown in (a)-(c).

[0026] Step 1.1: Count the maximum and minimum values ​​of three-dimensional discrete points in the X, Y, and Z directions. The specific process is as follows: (1) Count the maximum and minimum values ​​minx, minY, minZ, maxX, maxY, maxZ in the X, Y, and Z directions respectively.

[0027] (2) Construct a KD tree and use existing algorithms such as octree statistics and grid statistics to calculate the point cloud density.

[0028] (3) Use dx, dy, and dz as intervals to slice the disordered three-dimensional discrete points from small to large in the corresponding directions of X, Y, and Z.

[0029] in, dx, dy, dz The calculation formula is as follows:

[0030] Step 2: Reduce the dimension of the three-dimensional slices according to the direction of the slices so that the points in the slices are located in the same plane.

[0031] Step 2.1, establish a plane coordinate system for the slice with two coordinate axes (such as Y and Z axes) perpendicular to the slice direction, and set the slice direction coordinates to 0. That is, reduce the 3D point cloud to a 2D point cloud, and only consider the plane coordinates of the slice.

[0032] Step 3, set the threshold for all 2D slice data in order, and find the concave hull on its slice surface, that is, the minimum concave polygon containing all points, and mark the points contained in the outer boundary of the minimum concave polygon, such as Figure 3 shown.

[0033] Step 3.1, use the classic rolling ball method to find the concave hull of each section. The specific process is as follows: (1) Find the polar center P0 of the two-dimensional point set; (2) Find the leftmost point, which is the first point P1 of the concave edge. (3) The radius of the ball is set by the rolling ball method, with P1 as the center, and the point set P within the fixed radius is selected. nearest ; (4) Traverse P nearest , taking the angle P0->P1 as the reference, rotate counterclockwise to find the point with the smallest counterclockwise angle and add it to the concave hull; (5) Repeat the above steps until no new points are found or a point that has been previously added to the concave hull is encountered.

[0034] Step 3.2, for the points on the concave hull obtained in step 3.1, find their subscripts in the original unordered three-dimensional discrete point cloud through the KD tree and mark them.

[0035] Step 4: Process the Y and Z directions according to the above steps, and count the points that are repeatedly marked three times to obtain a set of repeatedly marked points.

[0036] Step 4.1, by means of a counter, in the slices in the X, Y, and Z directions, the counter of the marked point is +1, and the repeatedly marked point is defined as being marked in all three directions of X, Y, and Z, that is, the counter is 3. In particular, a repeatedly marked point set can be obtained.

[0037] Step 5: Perform voxel filtering on the repeatedly marked point set to obtain the final feature points. Figure 4 shown.

[0038] In step 5.1, the repeatedly marked point set is voxelized with a voxel radius d to obtain a set of feature points with consistent resolution and uniform distribution. The point cloud within the voxel is downsampled to 1, which can effectively avoid the situation where the feature points are densely packed.

[0039] After completing the above 5 steps, the feature extraction of disordered 3D discrete point cloud can be completed. This process is very similar to CT slices in medicine. The point cloud is sliced, and then the possible feature points are marked according to the edge of the smallest concave polygon on the slice. The feature points are confirmed by repeated marking, and finally the feature points are avoided by voxel filtering. Compared with some classic feature point extraction algorithms such as SIFT, Harris3D, and NARF, this method is simpler and more direct in algorithm implementation. It does not need to calculate the gradient and normal direction of the point cloud data, but only determines the feature points by counting the concave edge points of each slice, and the processing efficiency is high.

[0040] During specific implementation, the above process can be automatically executed using computer software technology.

[0041] The implementation basis of each embodiment of the present invention is to implement programmed processing through a device with a processor function. Therefore, in engineering practice, the technical solutions and functions of each embodiment of the present invention are encapsulated into various modules. Based on this reality, on the basis of the above embodiments, an embodiment of the present invention provides a disordered three-dimensional discrete point cloud slicing feature extraction system, which is used to execute a disordered three-dimensional discrete point cloud slicing feature extraction method in the above method embodiment.

[0042] See also Figure 5 The system includes: a first main module, which is used to obtain the effective range and density of disordered three-dimensional discrete points, and set the interval of slices in the three directions of X, Y, and Z according to the density; a second main module, which is used to slice the disordered three-dimensional discrete points from small to large in the three directions of X, Y, and Z according to the set slice interval, and obtain two-dimensional slices in each direction; a third main module, which is used to calculate the concave hull of each section in the three directions of X, Y, and Z according to the two-dimensional slice data in each direction, and mark the points contained in the outer boundary of the concave hull; a fourth main module, which is used to count the repeatedly marked points to form a repeatedly marked point set; a fifth main module, which is used to perform voxel filtering on the repeatedly marked point set to obtain extracted feature points.

[0043] The embodiment of the present invention provides a slicing feature extraction system for disordered 3D discrete point clouds, which solves the problem that existing disordered 3D discrete point cloud feature extraction algorithms are complex and time-consuming to process. Figure 5Several modules in it perform slicing operations and dimensionality reduction on the x, y, and z directions of the disordered 3D discrete points, find the smallest concave polygon on the slice and mark the edge points. Finally, the point set that has been marked multiple times is voxel filtered to obtain the final feature points.

[0044] It should be noted that the system embodiment provided by the present invention is used to implement the methods in the above method embodiment as well as the methods in other method embodiments provided by the present invention. The only difference is that the corresponding functional modules are set. The principle is basically the same as the principle of the above system embodiment provided by the present invention. As long as the technical personnel in this field refer to the specific technical solutions in other method embodiments on the basis of the above system embodiment, obtain the corresponding technical means and the technical solutions composed of these technical means by combining the technical features, and on the premise of ensuring the practicality of the technical solutions, improve the modules in the above system embodiment to obtain the corresponding system class embodiments, which are used to implement the methods in other method class embodiments. For example: Based on the content of the above system embodiment, as a preferred embodiment, in an unordered three-dimensional discrete point cloud slicing feature extraction system provided in an embodiment of the present invention, the second main module is further used to execute the following instructions: Count the maximum and minimum values ​​of disordered three-dimensional discrete points in the X, Y, and Z directions; Use the intervals dx, dy, and dz to slice the disordered three-dimensional discrete points from small to large in the corresponding directions of X, Y, and Z.

[0045] Based on the content of the above system embodiment, as a preferred embodiment, in an unordered three-dimensional discrete point cloud slicing feature extraction system provided in an embodiment of the present invention, the second main module is further used to execute the following instructions: A plane coordinate system is established for the section with two coordinate axes perpendicular to the slicing direction, and the slicing direction coordinate is set to 0 to reduce the 3D point cloud to a 2D point cloud.

[0046] Based on the content of the above system embodiment, as a preferred embodiment, in an unordered three-dimensional discrete point cloud slicing feature extraction system provided in an embodiment of the present invention, the third main module is further used to execute the following instructions: Use the classic rolling ball method to find the concave hull of each section; For the points on the concave hull obtained, the subscripts in the original unordered three-dimensional discrete point cloud are found through the KD tree and marked.

[0047] Based on the content of the above system embodiment, as a preferred embodiment, in an unordered three-dimensional discrete point cloud slicing feature extraction system provided in an embodiment of the present invention, the third main module is further used to execute the following instructions: Find the polar center P0 of the two-dimensional point set; Find the leftmost point as the first point P1 of the concave edge; The radius of the ball set by the rolling ball method is used to filter out the point set Pnearest within the fixed radius with P1 as the center; Traverse Pnearest, take the angle P0->P1 as the reference, rotate counterclockwise to find the point with the smallest counterclockwise angle and add it to the concave hull; Repeat the above steps until no new points are found or a point that has been previously added to the concave hull is encountered.

[0048] Based on the content of the above system embodiment, as a preferred embodiment, in the embodiment of the present invention, a disordered three-dimensional discrete point cloud slicing feature extraction system is provided, wherein the fifth main module is further used to execute the following instructions: The repeatedly marked point set is voxelized with voxel radius to obtain a set of feature points with consistent resolution and uniform distribution.

[0049] Based on the same inventive concept as the above embodiment, the present invention further provides an electronic device, including a processor and a memory; the memory stores program instructions executed by the processor, and the processor calls the program instructions to execute the steps of the unordered three-dimensional discrete point cloud slicing feature extraction method as follows: Get the effective range and density of disordered 3D discrete points, and set the interval of slices in X, Y, and Z directions according to the density; According to the set slice interval, the disordered three-dimensional discrete points are sliced ​​from small to large in the three directions of X, Y, and Z to obtain two-dimensional slices in each direction; According to the two-dimensional slice data in each direction, the concave hull of each slice in the X, Y, and Z directions is calculated respectively, and the points contained in the outer boundary of the concave hull are marked; Count the repeatedly marked points to form a repeatedly marked point set; Voxel filtering is performed on the repeatedly marked point set to obtain the extracted feature points.

[0050] Based on the same inventive concept as the above embodiment, the present invention further provides a non-transitory computer-readable storage medium, wherein the non-transitory computer-readable storage medium stores computer instructions, wherein the computer instructions enable the computer to execute the steps of the method for extracting features from a slicing-based unordered three-dimensional discrete point cloud: Get the effective range and density of disordered 3D discrete points, and set the interval of slices in X, Y, and Z directions according to the density; According to the set slice interval, the disordered three-dimensional discrete points are sliced ​​from small to large in the three directions of X, Y, and Z to obtain two-dimensional slices in each direction; According to the two-dimensional slice data in each direction, the concave hull of each slice in the X, Y, and Z directions is calculated respectively, and the points contained in the outer boundary of the concave hull are marked; Count the repeatedly marked points to form a repeatedly marked point set; Voxel filtering is performed on the repeatedly marked point set to obtain the extracted feature points.

[0051] In summary, the present invention is inspired by CT slices in medicine and proposes a slicing feature extraction algorithm for disordered three-dimensional discrete point clouds. First, slice operations are performed on the x, y, and z directions of disordered three-dimensional discrete points and dimensionality reduction is performed. Then, the minimum concave polygon is found on the slice and the edge points are marked. Finally, the point set that has been marked multiple times is voxel filtered to obtain the final feature points. The method of the present invention has a simple processing process and high computational efficiency. It has no requirements for data points and can process three-dimensional entities of any shape and size. It is particularly suitable for fast feature extraction of three-dimensional geographic information data. In addition, the method of the present invention can avoid complex three-dimensional spatial feature calculations. The calculations within each slice are simple, and the large number of independent subtasks formed by slicing are also more conducive to parallel processing, which can stably describe the point cloud.

[0052] 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 it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the technical solutions of the embodiments of the present invention.

Claims

1. A slicing feature extraction method for disordered three-dimensional discrete point clouds, characterized in that: include: Get the effective range and density of disordered 3D discrete points, and set the interval of slices in X, Y, and Z directions according to the density; According to the set slice interval, the disordered three-dimensional discrete points are sliced ​​from small to large in the three directions of X, Y, and Z to obtain two-dimensional slices in each direction; According to the two-dimensional slice data in each direction, the concave hull of each slice in the X, Y, and Z directions is calculated respectively, and the points contained in the outer boundary of the concave hull are marked; Count the repeatedly marked points to form a repeatedly marked point set; Voxel filtering is performed on the repeatedly marked point set to obtain the extracted feature points.

2. According to claim 1, a slicing feature extraction method for disordered three-dimensional discrete point clouds is characterized in that: Slice the disordered 3D discrete points from small to large in the X, Y, and Z directions, including: Count the maximum and minimum values ​​of disordered three-dimensional discrete points in the X, Y, and Z directions; Use the intervals dx, dy, and dz to slice the disordered three-dimensional discrete points from small to large in the corresponding directions of X, Y, and Z.

3. According to claim 2, a slicing feature extraction method for disordered three-dimensional discrete point clouds is characterized in that: Get 2D slices in all directions, including: A plane coordinate system is established for the section with two coordinate axes perpendicular to the slicing direction, and the slicing direction coordinate is set to 0 to reduce the 3D point cloud to a 2D point cloud.

4. The method for extracting features from unordered three-dimensional discrete point clouds by slicing according to claim 1, characterized in that: Calculate the concave hull of each section in the X, Y, and Z directions separately, and mark the points contained in the outer boundary of the concave hull, including: Use the classic rolling ball method to find the concave hull of each section; For the points on the concave hull obtained, the subscripts in the original unordered three-dimensional discrete point cloud are found through the KD tree and marked.

5. The method for extracting features by slicing unordered three-dimensional discrete point clouds according to claim 4, characterized in that: Use the classic rolling ball method to find the concave hull of each section, including: Find the polar center P0 of the two-dimensional point set; Find the leftmost point as the first point P1 of the concave edge; The radius of the ball set by the rolling ball method is P1 as the center, and the point set P within the fixed radius is screened out. nearest ; Traverse P nearest , taking the angle P0->P1 as the reference, rotate counterclockwise to find the point with the smallest counterclockwise angle and add it to the concave hull; Repeat the above steps until no new points are found or a point that has been previously added to the concave hull is encountered.

6. The method for extracting features from an unordered three-dimensional discrete point cloud by slicing according to claim 1, characterized in that: Voxel filtering is performed on the repeatedly labeled point set, including: The repeatedly marked point set is voxelized with voxel radius to obtain a set of feature points with consistent resolution and uniform distribution.

7. A slicing feature extraction system for disordered three-dimensional discrete point clouds, characterized in that: include: The first main module is used to obtain the effective range and density of disordered three-dimensional discrete points, and set the interval of slices in the X, Y and Z directions according to the density; The second main module is used to slice the disordered three-dimensional discrete points from small to large in the three directions of X, Y and Z according to the set slice interval to obtain two-dimensional slices in each direction; The third main module is used to calculate the concave hull of each section in the X, Y and Z directions according to the two-dimensional slice data in each direction, and mark the points contained in the outer boundary of the concave hull; The fourth main module is used to count the repeatedly marked points to form a repeatedly marked point set; The fifth main module is used to perform voxel filtering on the repeatedly marked point set to obtain extracted feature points.

8. An electronic device, characterized in that: It includes a processor and a memory; the memory stores program instructions executed by the processor, and the processor calls the program instructions to execute the steps of a slicing feature extraction method for an unordered three-dimensional discrete point cloud as described in any one of claims 1 to 6.

9. A non-transitory computer readable storage medium, characterized in that: The non-transitory computer-readable storage medium stores computer instructions, and the computer instructions enable the computer to execute the steps of a slicing feature extraction method for an unordered three-dimensional discrete point cloud as described in any one of claims 1 to 6.

Images