Method and apparatus for constructing point cloud structural feature descriptors

Abstract

This disclosure provides a method and apparatus for constructing a point cloud structural feature descriptor. The method includes: determining feature information of m shape structures; determining one shape structure as the nearest neighbor structure of another shape structure based on the center point coordinates of the m shape structures and a preset rule; and constructing a structural feature descriptor based on the feature information of the two nearest neighbor shape structures. The structural feature descriptor constructed by this disclosure intuitively and vividly represents the information of the shape structure itself and its relative positional relationship with its nearest neighbor shape structure. Furthermore, the structural feature descriptor has a low dimensionality, resulting in less computation time during large-scale point cloud registration and effectively improving computational efficiency.

Description

Technical Field

[0001] This disclosure relates to the field of data processing technology, specifically to a method and apparatus for constructing point cloud structural feature descriptors. Background Technology

[0002] Point cloud registration is the process of converting point cloud data with independent coordinate systems obtained by scanning the object under test multiple times from multiple perspectives to a global coordinate system and then stitching them together.

[0003] In related technologies, point cloud registration relies on local geometric descriptors to establish the correspondence between point clouds. Most geometric descriptors are in floating-point form, and Euclidean distance needs to be calculated in a high-dimensional vector space for registration. When the amount of point cloud data is large, the calculation process is very time-consuming. Summary of the Invention

[0004] The main objective of this disclosure is to provide a method and apparatus for constructing point cloud structural feature descriptors, thereby solving the problem of high computational load in the point cloud registration process.

[0005] To achieve the above objectives, a first aspect of this disclosure provides a method for constructing a point cloud structural feature descriptor, comprising:

[0006] Determine the feature information of m shape structures. The feature information includes shape structure labels, vertex coordinate set in the original coordinate system, center point coordinates, and direction lines. The m shape structures are obtained by segmenting the same point cloud data. Based on the center point coordinates of m shape structures and a preset rule, one shape structure among the m shape structures is determined as the nearest neighbor structure of another shape structure. The preset rule includes that the directional lines of the two shape structures are not parallel. Based on the feature information of the two nearest neighbor shapes, the following structural feature descriptor is constructed:

[0007] in, For the label of the i-th shape structure, The center point of the i-th shape structure and the center point of the j-th shape structure European distance, The directional straight line of the i-th shape structure and the direction of the j-th shape structure The included angle, The directional straight line of the i-th shape structure and the direction of the j-th shape structure The shortest distance between them Let be the length of the longest side in the OBB bounding box of the i-th shape structure. Let be the length of the second longest side in the OBB bounding box of the i-th shape structure. Let be the length of the shortest side in the OBB bounding box of the i-th shape structure.

[0008] Optionally, the shape structure label of the i-th shape structure is determined, including: If the shape structure is planar, then the shape structure label ; If the shape structure is spherical, then the shape structure label ; If the shape structure is a cylindrical surface, then the shape structure label .

[0009] Optionally, determining the coordinates of the center point of the i-th shape structure includes: Let the point set corresponding to the i-th shape structure be... as follows:

[0010] Then point set center point as follows:

[0011] in, The point set corresponding to the i-th shape structure The k-th point, For point coordinates Let be the total number of data points for the i-th shape structuring element. Let i be the center point of the point set corresponding to the i-th shape structure. Center point The coordinates.

[0012] Optionally, the set of vertex coordinates of the i-th shape structure in the original coordinate system is determined, including: Based on point set The expression that determines the covariance matrix A of its data point distribution is as follows:

[0013] Diagonalize the covariance matrix A, i.e.:

[0014] in,

[0015] Three feature vectors The directions are used as the three coordinate axes of the minimum bounding box to establish a new coordinate system; Point set Transform from the original coordinate system to the new coordinate system to obtain the point set in the new coordinate system. :

[0016] Set of points The maximum and minimum values ​​of the coordinates on the x, y, and z axes of the new coordinate system are respectively and , and , and ; Based on this, we can obtain the set of vertex coordinates of the bounding box in the new coordinate system. :

[0017] Using matrix Q, the vertices are transformed from the new coordinate system to the original coordinate system, resulting in the set of vertex coordinates in the original coordinate system. :

[0018] Where S is a diagonal matrix with eigenvalues ​​as diagonal elements. The normalized unit eigenvector corresponding to the eigenvalue. The set of points corresponding to the i-th shape structure in the new coordinate system The k-th point, For point coordinates Let be the total number of data points for the i-th shape structuring element in the new coordinate system. Let these be the vertex coordinates of the bounding box in the new coordinate system. Let be the vertex coordinates of the i-th shape structuring element in the original coordinate system.

[0019] Optionally, determining the directional line of the i-th shape structure includes: Let REC1 and REC2 be the pair of relative rectangles with the largest area in the bounding box of the i-th shape structure; The four vertices that make up REC1 are respectively , , , ; Calculate the normal vector of REC1 using the following formula. for:

[0020] Therefore, a path passing through the center point of the i-th shape structure can be determined according to the following formula. And the direction vector is The straight line is used as the direction line of the i-th shape structure:

[0021] in, R is a real number.

[0022] Optionally, determining one shape structure as the nearest neighbor structure of another shape structure based on the center point coordinates of the m shape structures and a preset rule includes: A KD tree is constructed based on the center points of m shape structures; The nearest neighbor shape structure for each of the m shape structures is determined based on the established KD tree. Based on the nearest neighbor structure of each shape and preset rules, determine one shape structure among m shape structures as the nearest neighbor structure of another shape structure.

[0023] A second aspect of this disclosure provides a point cloud structural feature descriptor construction apparatus, comprising: The feature determination module is used to determine the feature information of m shape structures. The feature information includes shape structure labels, vertex coordinate set in the original coordinate system, center point coordinates, and direction lines. The m shape structures are obtained by segmenting the same point cloud data. The structure determination module is used to determine one of the m shape structures as the nearest neighbor structure of another shape structure based on the center point coordinates of the m shape structures and preset rules. The preset rules include that the directional straight lines of the two shape structures are not parallel. The descriptor building module is used to construct structural feature descriptors based on the feature information of the two nearest neighboring shape structures, as follows:

[0024] in, For the label of the i-th shape structure, The center point of the i-th shape structure and the center point of the j-th shape structure European distance, The directional straight line of the i-th shape structure and the direction of the j-th shape structure The included angle, The directional straight line of the i-th shape structure and the direction of the j-th shape structure The shortest distance between them Let be the length of the longest side in the OBB bounding box of the i-th shape structure. Let be the length of the second longest side in the OBB bounding box of the i-th shape structure. Let be the length of the shortest side in the OBB bounding box of the i-th shape structure.

[0025] Optionally, the structure determination module is specifically used for: A KD tree is constructed based on the center points of m shape structures; The nearest neighbor shape structure for each of the m shape structures is determined based on the established KD tree. Based on the nearest neighbor structure of each shape and preset rules, determine one shape structure among m shape structures as the nearest neighbor structure of another shape structure.

[0026] A third aspect of this disclosure provides an electronic device comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores a computer program executable by the at least one processor, the computer program being executed by the at least one processor to cause the at least one processor to perform the point cloud structural feature descriptor construction method according to any one of the first aspects.

[0027] A fourth aspect of this disclosure provides a computer-readable storage medium storing computer instructions for causing a computer to perform the point cloud structural feature descriptor construction method according to any one of the first aspects.

[0028] The point cloud structure feature descriptor construction method and apparatus provided in this disclosure determine the structure feature descriptor of each shape structure in the point cloud data by using the feature information of m shape structures obtained by segmenting the same point cloud data and preset rules. This allows point cloud registration to be performed based on the structure feature descriptor, thereby solving the problem of large computational load in the point cloud registration process.

[0029] It should be understood that the description in the Summary of the Invention section is not intended to limit the key or essential features of the embodiments of this disclosure, nor is it intended to restrict the scope of this disclosure. Other features of this disclosure will become readily apparent from the following description. Attached Figure Description

[0030] To more clearly illustrate the technical solutions in the specific embodiments of this disclosure or the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this disclosure. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0031] Figure 1 A flowchart illustrating the point cloud structural feature descriptor construction method provided in this embodiment of the disclosure; Figure 2 A block diagram illustrating a point cloud structure feature descriptor construction apparatus provided in an embodiment of this disclosure; Figure 3 A block diagram of an electronic device provided in an embodiment of this disclosure. Detailed Implementation

[0032] To enable those skilled in the art to better understand the present disclosure, the technical solutions of the present disclosure will be clearly and completely described below with reference to the accompanying drawings of the embodiments. Obviously, the described embodiments are only some embodiments of the present disclosure, and not all embodiments. Based on the embodiments of the present disclosure, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present disclosure.

[0033] It should be noted that the terms "comprising" and "having" and any variations thereof in the specification, claims and accompanying drawings of this disclosure are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are explicitly listed, but may include other steps or units that are not explicitly listed or that are inherent to such process, method, product or device.

[0034] It should be noted that, unless otherwise specified, the embodiments and features described in this disclosure can be combined with each other. This disclosure will now be described in detail with reference to the accompanying drawings and embodiments.

[0035] Figure 1 This is a flowchart illustrating the point cloud structural feature descriptor construction method provided in this embodiment of the disclosure. See also... Figure 1 The method includes the following steps: Step 102: Determine the feature information of m shape structures.

[0036] Among them, shape structure refers to multiple sub-shape structure elements formed after the point cloud is segmented into shapes. The types of multiple sub-shape structure elements can be the same or different.

[0037] For example, objects in real-world scenarios can typically be represented as a collection of planes, spheres, and cylinders; therefore, the shape structure can be a plane, a sphere, or a cylinder.

[0038] In some alternative implementations, m shape structures can be obtained from the same point cloud through shape segmentation.

[0039] Each shape structure's feature information includes a shape structure label, a set of vertex coordinates in the original coordinate system, center point coordinates, and a direction line.

[0040] The following example, using the determination of the feature information of the i-th shape structure, illustrates how to determine the feature information of m shape structures. It can be understood that i ∈ m, and m is a positive integer greater than or equal to 2.

[0041] (1) Determine the shape structure label of the i-th shape structure. As mentioned above, objects in real-world scenarios can typically be represented as a collection of planes, spheres, and cylinders. Therefore, in this embodiment of the disclosure, the shape structure label can be a plane label, a sphere label, and a cylinder label.

[0042] For example, when the shape structure is planar, the shape structure label .

[0043] For example, when the shape structure is spherical, the shape structure label .

[0044] For example, when the shape structure is a cylindrical surface, the shape structure label... .

[0045] It should be noted that the label representation method in the above example is only an example. Other representation methods are also possible, as long as they can distinguish whether the shape structure is a plane, a sphere, or a cylinder. These will not be elaborated here.

[0046] (2) Determine the coordinates of the center point of the i-th shape structure. Let the point set corresponding to the i-th shape structure be... as follows:

[0047] Then point set center point as follows:

[0048] in, The point set corresponding to the i-th shape structure The k-th point, For point coordinates Let be the total number of data points for the i-th shape structuring element. Let i be the center point of the point set corresponding to the i-th shape structure. Center point The coordinates.

[0049] (3) Determine the set of vertex coordinates of the i-th shape structure in the original coordinate system. Based on point set The expression that determines the covariance matrix A of its data point distribution is as follows:

[0050] Diagonalize the covariance matrix A, i.e.:

[0051] in,

[0052] Three feature vectors The directions are used as the three coordinate axes of the minimum bounding box to establish a new coordinate system; Point set Transform from the original coordinate system to the new coordinate system using the following formula:

[0053] Obtain the point set in the new coordinate system :

[0054] Set of points The maximum and minimum values ​​of the coordinates on the x, y, and z axes of the new coordinate system are respectively and , and , and ; Based on this, we can obtain the set of vertex coordinates of the bounding box in the new coordinate system. :

[0055] Using matrix Q, the vertices are transformed from the new coordinate system to the original coordinate system using the following formula:

[0056] Obtain the set of vertex coordinates in the original coordinate system :

[0057] Where S is a diagonal matrix with eigenvalues ​​as diagonal elements. The normalized unit eigenvector corresponding to the eigenvalue. The set of points corresponding to the i-th shape structure in the new coordinate system The k-th point, For point coordinates Let be the total number of data points for the i-th shape structuring element in the new coordinate system. Let these be the vertex coordinates of the bounding box in the new coordinate system. Let be the vertex coordinates of the i-th shape structuring element in the original coordinate system.

[0058] (4) Determine the direction line of the i-th shape structure. Let REC1 and REC2 be the pair of relative rectangles with the largest area in the bounding box of the i-th shape structure; The four vertices that make up REC1 are respectively , , , ; Calculate the normal vector of REC1 using the following formula. for:

[0059] Therefore, a path passing through the center point of the i-th shape structure can be determined according to the following formula. And the direction vector is The straight line is used as the direction line of the i-th shape structure:

[0060] in, R is a real number.

[0061] Step 104: Based on the center point coordinates of the m shape structures and the preset rules, determine one of the m shape structures as the nearest neighbor structure of another shape structure.

[0062] The preset rules include that the directional straight lines of the two shapes are not parallel.

[0063] Specifically, determining one shape structure among m shape structures as the nearest neighbor structure of another shape structure may include the following steps: 1042 (not shown in the figure), 1044 (not shown in the figure), and 1046 (not shown in the figure). Wherein: Step 1042: Build a KD tree based on the center points of m shape structures.

[0064] Step 1044: Determine the nearest neighbor shape structure for each of the m shape structures based on the established KD tree.

[0065] Step 1046: Determine one of the m shape structures as the nearest neighbor structure of another shape structure based on the nearest neighbor structure of each shape and the preset rules.

[0066] For example, the center points of all m shaped structures are:

[0067] in, Let d be the center point of the d-th shape among m shapes.

[0068] Sort all center points according to their x-coordinate values ​​and select the median value. As the root node, all nodes less than The point is in the left subtree, all points greater than The point is in the right subtree.

[0069] The points on the left and right subtrees are then sorted according to their y-coordinate values. The root node of the subtree is found, and the subtree is generated again. Then, the points in the subtree are sorted according to their z-coordinate values, and then according to their x-coordinate values, and so on, in a loop, so that each center point is assigned to a different subtree.

[0070] By searching a KD tree, we can quickly find the nearest center point of a center point. Combined with the fact that the directional lines of two shapes are not parallel, we can quickly determine the nearest neighbor shape of a shape.

[0071] Euclidean distance can be used to measure the distance between two center points.

[0072] Step 106: Based on the feature information of the two nearest neighboring shapes, construct the following structural feature descriptor:

[0073] in, For the label of the i-th shape structure, The center point of the i-th shape structure and the center point of the j-th shape structure European distance, The directional straight line of the i-th shape structure and the direction of the j-th shape structure The included angle, The directional straight line of the i-th shape structure and the direction of the j-th shape structure The shortest distance between them Let be the length of the longest side in the OBB bounding box of the i-th shape structure. Let be the length of the second longest side in the OBB bounding box of the i-th shape structure. Let be the length of the shortest side in the OBB bounding box of the i-th shape structure.

[0074] It should be noted that the directional straight line of the i-th shape structure and the direction of the j-th shape structure The included angle is an acute angle.

[0075] The point cloud structure feature descriptor construction method provided in this embodiment determines the feature information of m shape structures, and based on the center point coordinates of the m shape structures and preset rules, determines one shape structure as the nearest neighbor structure of another shape structure. Then, a structure feature descriptor is constructed based on the feature information of the two nearest neighbor shape structures. The structure feature descriptor includes the label of the shape structure, the Euclidean distance between the center points of the two shape structures, the angle and shortest distance between the directional lines of the two shape structures, and the lengths of the longest, second longest, and shortest sides in the OBB bounding box of the shape structure. This makes the structure feature descriptor more intuitive and vivid in expressing the information of the shape structure itself and its relative positional relationship with its nearest neighbor shape structure. Moreover, the structure feature descriptor has a low dimension, which reduces the computation time in large-scale point cloud registration and can effectively improve computational efficiency.

[0076] It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases the steps shown or described may be executed in a different order than that shown here.

[0077] The above is an introduction to the method embodiments. The following describes the solution described in this disclosure further through device embodiments.

[0078] Figure 2 A block diagram illustrating a point cloud structural feature descriptor construction apparatus provided in an embodiment of this disclosure. See also... Figure 2 The device includes: Feature determination module 202 is used to determine the feature information of m shape structures.

[0079] The feature information includes shape structure labels, vertex coordinate set in the original coordinate system, center point coordinates, and direction lines. The m shape structures are obtained by segmenting the same point cloud data.

[0080] The structure determination module 204 is used to determine one of the m shape structures as the nearest neighbor structure of another shape structure based on the center point coordinates of the m shape structures and preset rules.

[0081] The preset rule includes that the directional straight lines of the two shape structures are not parallel.

[0082] Descriptor building module 206 is used to construct the following structural feature descriptor based on the feature information of the two nearest neighboring shape structures:

[0083] in, For the label of the i-th shape structure, The center point of the i-th shape structure and the center point of the j-th shape structure European distance, The directional straight line of the i-th shape structure and the direction of the j-th shape structure The included angle, The directional straight line of the i-th shape structure and the direction of the j-th shape structure The shortest distance between them Let be the length of the longest side in the OBB bounding box of the i-th shape structure. Let be the length of the second longest side in the OBB bounding box of the i-th shape structure. Let be the length of the shortest side in the OBB bounding box of the i-th shape structure.

[0084] The point cloud structure feature descriptor construction device provided in this embodiment determines the feature information of m shape structures, and based on the center point coordinates of the m shape structures and preset rules, determines one shape structure among the m shape structures as the nearest neighbor structure of another shape structure. Then, it constructs a structure feature descriptor based on the feature information of the two nearest neighbor shape structures. The structure feature descriptor includes the label of the shape structure, the Euclidean distance between the center points of the two shape structures, the angle and shortest distance between the directional lines of the two shape structures, and the lengths of the longest side, the second longest side, and the shortest side in the OBB bounding box of the shape structure. This makes the structure feature descriptor more intuitive and vivid in expressing the information of the shape structure itself and its relative positional relationship with its nearest neighbor shape structure. Moreover, the dimension of the structure feature descriptor is low, and the calculation time is less in large-scale point cloud registration, which can effectively improve the calculation efficiency.

[0085] In some alternative embodiments, the feature determination module 202 is specifically used for: Define the shape structure label when the shape structure is planar. ; Given a spherical shape, determine the shape structure label. ; Define the shape structure label for a cylindrical surface. .

[0086] In some optional embodiments, the feature determination module 202 is specifically used to determine the coordinates of the center point of the i-th shape structure: Let the point set corresponding to the i-th shape structure be... as follows:

[0087] Then point set center point as follows:

[0088] in, The point set corresponding to the i-th shape structure The k-th point, For point coordinates Let be the total number of data points for the i-th shape structuring element. Let i be the center point of the point set corresponding to the i-th shape structure. Center point The coordinates.

[0089] In some alternative embodiments, the feature determination module 202 is specifically used to determine the set of vertex coordinates of the i-th shape structure in the original coordinate system: Based on point set The expression that determines the covariance matrix A of its data point distribution is as follows:

[0090] Diagonalize the covariance matrix A, i.e.:

[0091] in,

[0092] Three feature vectors The directions are used as the three coordinate axes of the minimum bounding box to establish a new coordinate system; Point set Transform from the original coordinate system to the new coordinate system to obtain the point set in the new coordinate system. :

[0093] Set of points The maximum and minimum values ​​of the coordinates on the x, y, and z axes of the new coordinate system are respectively and , and , and ; Based on this, we can obtain the set of vertex coordinates of the bounding box in the new coordinate system. :

[0094] Using matrix Q, the vertices are transformed from the new coordinate system to the original coordinate system, resulting in the set of vertex coordinates in the original coordinate system. :

[0095] Where S is a diagonal matrix with eigenvalues ​​as diagonal elements. The normalized unit eigenvector corresponding to the eigenvalue. The set of points corresponding to the i-th shape structure in the new coordinate system The k-th point, For point coordinates Let be the total number of data points for the i-th shape structuring element in the new coordinate system. Let these be the vertex coordinates of the bounding box in the new coordinate system. Let be the vertex coordinates of the i-th shape structuring element in the original coordinate system.

[0096] In some alternative embodiments, the feature determination module 202 is specifically used to determine the directional straight line of the i-th shape structure: Let REC1 and REC2 be the pair of relative rectangles with the largest area in the bounding box of the i-th shape structure; The four vertices that make up REC1 are respectively , , , ; Calculate the normal vector of REC1 using the following formula. for:

[0097] Therefore, a path passing through the center point of the i-th shape structure can be determined according to the following formula. And the direction vector is The straight line is used as the direction line of the i-th shape structure:

[0098] in, R is a real number.

[0099] In some optional embodiments, the structure determination module 204 is specifically used to: establish a KD tree based on the center points of m shape structures; determine the nearest neighbor shape structure of each of the m shape structures according to the established KD tree; and determine one shape structure among the m shape structures as the nearest neighbor structure of another shape structure according to the nearest neighbor shape structure of each shape and a preset rule.

[0100] The specific methods of execution of each unit in the above device embodiments have been described in detail in the embodiments related to the method, and will not be elaborated here.

[0101] This disclosure also provides an electronic device, such as... Figure 3 As shown, the electronic device includes one or more processors 301 and a memory 302. Figure 3 Take processor 301 as an example.

[0102] The controller may also include an input device 303 and an output device 304.

[0103] The processor 301, memory 302, input device 303, and output device 304 can be connected via a bus or other means. Figure 3 Taking the example of a connection between China and Israel via a bus.

[0104] Processor 301 can be a central processing unit (CPU), or it can be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, or combinations of the above types of chips. The general-purpose processor can be a microprocessor or any conventional processor.

[0105] The memory 302, as a non-transitory computer-readable storage medium, can be used to store non-transitory software programs, non-transitory computer-executable programs, and modules, such as the program instructions / modules corresponding to the control method in this embodiment. The processor 301 executes various functional applications and data processing of the server by running the non-transitory software programs, instructions, and modules stored in the memory 302, thereby implementing the point cloud structure feature descriptor construction method of the above method embodiment.

[0106] Memory 302 may include a program storage area and a data storage area. The program storage area may store the operating system and applications required for at least one function; the data storage area may store data created by the use of the processing device operated by the server. Furthermore, memory 302 may include high-speed random access memory and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some embodiments, memory 302 may optionally include memory remotely located relative to processor 301, and these remote memories can be connected to a network connection device via a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.

[0107] Input device 303 can receive input numeric or character information, and generate key signal inputs related to user settings and function control of the server's processing device. Output device 304 may include display devices such as a display screen.

[0108] One or more modules are stored in memory 302, and when executed by one or more processors 301, they perform actions such as... Figure 1 The method shown.

[0109] Those skilled in the art will understand that all or part of the processes in the above method embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, it can include the processes described in the above method embodiments. The storage medium can be a magnetic disk, optical disk, read-only memory (ROM), random access memory (RAM), flash memory (FM), hard disk drive (HDD), or solid-state drive (SSD), etc.; the storage medium can also include combinations of the above types of memory.

[0110] Although embodiments of the present disclosure have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the present disclosure, and such modifications and variations all fall within the scope defined by the appended claims.

Claims

1. A method for constructing a point cloud structural feature descriptor, characterized in that, include: Determine the feature information of m shape structures. The feature information includes shape structure labels, vertex coordinate set in the original coordinate system, center point coordinates, and direction lines. The m shape structures are obtained by segmenting the same point cloud data. Based on the center point coordinates of m shape structures and a preset rule, one shape structure among the m shape structures is determined as the nearest neighbor structure of another shape structure. The preset rule includes that the directional lines of the two shape structures are not parallel. Based on the feature information of the two nearest neighbor shapes, the following structural feature descriptor is constructed: in, For the label of the i-th shape structure, The center point of the i-th shape structure and the center point of the j-th shape structure European distance, The directional straight line of the i-th shape structure and the direction of the j-th shape structure The included angle, The directional straight line of the i-th shape structure and the direction of the j-th shape structure The shortest distance between them Let be the length of the longest side in the OBB bounding box of the i-th shape structure. Let be the length of the second longest side in the OBB bounding box of the i-th shape structure. Let be the length of the shortest side in the OBB bounding box of the i-th shape structure.

2. The method according to claim 1, characterized in that, Determine the shape structure label of the i-th shape structure, including: If the shape structure is planar, then the shape structure label ; If the shape structure is spherical, then the shape structure label ; If the shape structure is a cylindrical surface, then the shape structure label .

3. The method according to claim 2, characterized in that, Determining the coordinates of the center point of the i-th shape structure includes: Let the point set corresponding to the i-th shape structure be... as follows: Then point set center point as follows: in, The point set corresponding to the i-th shape structure The k-th point, For point coordinates Let be the total number of data points for the i-th shape structuring element. Let i be the center point of the point set corresponding to the i-th shape structure. Center point The coordinates.

4. The method according to claim 3, characterized in that, Determine the set of vertex coordinates of the i-th shape structure in the original coordinate system, including: Based on point set The expression that determines the covariance matrix A of its data point distribution is as follows: Diagonalize the covariance matrix A, i.e.: in, Three feature vectors The directions are used as the three coordinate axes of the minimum bounding box to establish a new coordinate system; Point set Transform from the original coordinate system to the new coordinate system to obtain the point set in the new coordinate system. : Set of points The maximum and minimum values ​​of the coordinates on the x, y, and z axes of the new coordinate system are respectively and , and , and ; Based on this, we can obtain the set of vertex coordinates of the bounding box in the new coordinate system. : Using matrix Q, the vertices are transformed from the new coordinate system to the original coordinate system, resulting in the set of vertex coordinates in the original coordinate system. : Where S is a diagonal matrix with eigenvalues ​​as diagonal elements. The normalized unit eigenvector corresponding to the eigenvalue. The set of points corresponding to the i-th shape structure in the new coordinate system The k-th point, For point coordinates Let be the total number of data points for the i-th shape structuring element in the new coordinate system. Let these be the vertex coordinates of the bounding box in the new coordinate system. Let be the vertex coordinates of the i-th shape structuring element in the original coordinate system.

5. The method according to claim 4, characterized in that, Determine the directional line of the i-th shape structure, including: Let REC1 and REC2 be the pair of relative rectangles with the largest area in the bounding box of the i-th shape structure; The four vertices that make up REC1 are respectively , , , ; Calculate the normal vector of REC1 using the following formula. for: Therefore, a path passing through the center point of the i-th shape structure can be determined according to the following formula. And the direction vector is The straight line is used as the direction line of the i-th shape structure: in, R is a real number.

6. The method according to claim 1, characterized in that, The process of determining the nearest neighbor structure of one of the m shape structures based on the center point coordinates of the m shape structures and preset rules includes: A KD tree is constructed based on the center points of m shape structures; The nearest neighbor shape structure for each of the m shape structures is determined based on the established KD tree. Based on the nearest neighbor structure of each shape and preset rules, determine one shape structure among m shape structures as the nearest neighbor structure of another shape structure.

7. A point cloud structural feature descriptor construction device, characterized in that, include: The feature determination module is used to determine the feature information of m shape structures. The feature information includes shape structure labels, vertex coordinate set in the original coordinate system, center point coordinates, and direction lines. The m shape structures are obtained by segmenting the same point cloud data. The structure determination module is used to determine one of the m shape structures as the nearest neighbor structure of another shape structure based on the center point coordinates of the m shape structures and preset rules. The preset rules include that the directional straight lines of the two shape structures are not parallel. The descriptor building module is used to construct structural feature descriptors based on the feature information of the two nearest neighboring shape structures, as follows: in, For the label of the i-th shape structure, The center point of the i-th shape structure and the center point of the j-th shape structure European distance, The directional straight line of the i-th shape structure and the direction of the j-th shape structure The included angle, The directional straight line of the i-th shape structure and the direction of the j-th shape structure The shortest distance between them Let be the length of the longest side in the OBB bounding box of the i-th shape structure. Let be the length of the second longest side in the OBB bounding box of the i-th shape structure. Let be the length of the shortest side in the OBB bounding box of the i-th shape structure.

8. The apparatus according to claim 7, characterized in that, The structure determination module is specifically used for: A KD tree is constructed based on the center points of m shape structures; The nearest neighbor shape structure for each of the m shape structures is determined based on the established KD tree. Based on the nearest neighbor structure of each shape and preset rules, determine one shape structure among m shape structures as the nearest neighbor structure of another shape structure.

9. An electronic device, characterized in that, The electronic device includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores a computer program executable by the at least one processor, the computer program being executed by the at least one processor to cause the at least one processor to perform the point cloud structure feature descriptor construction method according to any one of claims 1-6.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions for causing a computer to execute the point cloud structural feature descriptor construction method according to any one of claims 1-6.

Images