CN-121999253-A - Point cloud frame searching and matching method and device, electronic equipment and storage medium

CN121999253ACN 121999253 ACN121999253 ACN 121999253ACN-121999253-A

Abstract

The application provides a point cloud frame searching and matching method, a device, electronic equipment and a storage medium, wherein the method comprises the steps of determining first description sub-information of each point Yun Zhen in a point cloud frame sample set, and determining first hash coding information corresponding to each first description sub-information; determining second description sub-information of an input point cloud frame, determining second hash coding information corresponding to the second description sub-information, determining a similar candidate subset from a point cloud frame sample set according to the first hash coding information and the second hash coding information, wherein the similar candidate subset comprises M point cloud frames, M is a positive integer, and determining a target point cloud frame from the similar candidate subset according to the input point cloud frame. The application adds the descriptor hash search to screen, can avoid consuming a great deal of time by directly traversing the search descriptor information, thereby achieving the effect of greatly reducing the calculation amount of the descriptor-based point cloud frame search matching.

Inventors

GUO YUPENG
ZHANG LIANGLIANG
Peng Jili

Assignees

广州星程智能科技有限公司

Dates

Publication Date: 20260508
Application Date: 20251229

Claims (10)

1. A point cloud frame search matching method, the method comprising: determining first description sub-information of each point Yun Zhen in the point cloud frame sample set, and determining first hash coding information corresponding to each first description sub-information; determining second description sub-information of an input point cloud frame, and determining second hash coding information corresponding to the second description sub-information; according to the first hash coding information and the second hash coding information, determining a similar candidate subset from the point cloud frame sample set, wherein the similar candidate subset comprises M point cloud frames, and M is a positive integer; and determining a target point cloud frame from the similar candidate subset according to the input point cloud frame.
2. The method of claim 1, wherein determining the first descriptor information for each point Yun Zhen in the set of point cloud frame samples comprises: Performing two-dimensional projection dimension reduction processing on the point cloud frame aiming at any point cloud frame in the point cloud frame sample set to generate a first two-dimensional matrix; And generating first descriptor information of the point cloud frame according to the first two-dimensional matrix.
3. The method of claim 2, wherein determining the first hash encoded information corresponding to each of the first descriptor information comprises: And aiming at any point cloud frame in the point cloud frame sample set, carrying out hash coding on a first two-dimensional matrix in first description sub-information of the point cloud frame to obtain first hash coding information of the point cloud frame.
4. The method of claim 3, wherein hashing the two-dimensional matrix in the first descriptor information of the point cloud frame to obtain the first hash-coded information of the point cloud frame includes: Performing difference value downsampling on a first two-dimensional matrix in first description sub-information of the point cloud frame to obtain a second two-dimensional matrix; And performing discrete cosine transform on the second two-dimensional matrix to determine first hash coding information of the point cloud frame.
5. The method of claim 1, wherein said determining a target point cloud frame from said similar candidate subset from said input point cloud frame comprises: calculating the similarity between the input point cloud frame and each point cloud frame in the similar candidate subset; And taking the point cloud frame with the largest similarity in the similar candidate subset as the target point cloud frame.
6. The method of claim 5, wherein the calculating the similarity of the input point cloud frame to each point cloud frame in the similar candidate subset comprises: and calculating the similarity between the second descriptor information and the first descriptor information of each point Yun Zhen in the similar candidate subset.
7. The method of claim 6, wherein said calculating the similarity of the second descriptor information to the first descriptor information for each point Yun Zhen in the similar candidate subset comprises: determining a second phase map of the second descriptor information and a first phase map of the first descriptor information of each point Yun Zhen in the similar candidate subset; and determining the similarity according to the first phase map and the second phase map.
8. A point cloud frame search matching apparatus, the apparatus comprising: The first determining module is used for determining first description sub-information of each point Yun Zhen in the point cloud frame sample set and determining first hash coding information corresponding to each first description sub-information; The second determining module is used for determining second description sub-information of the input point cloud frame and determining second hash coding information corresponding to the second description sub-information; The screening module is used for determining a similar candidate subset from the point cloud frame sample set according to the first hash coding information and the second hash coding information, wherein the similar candidate subset comprises M point cloud frames, and M is a positive integer; And the searching matching module is used for determining target point cloud frames from the similar candidate subsets according to the input point cloud frames.
9. An electronic device comprising a processor, a memory and a computer program stored on the memory and executable on the processor, the computer program implementing the point cloud frame search matching method of any of claims 1 to 7 when executed by the processor.
10. A computer readable storage medium, characterized in that the computer readable storage medium has stored thereon a computer program which, when executed by a processor, implements the point cloud frame search matching method according to any of claims 1 to 7.

Description

Point cloud frame searching and matching method and device, electronic equipment and storage medium Technical Field The application belongs to the technical field of point cloud frame matching, and particularly relates to a point cloud frame searching and matching method, a device, electronic equipment and a storage medium. Background In the related technology, the application of the high-line number Lidar is wider and wider, and the traditional direct point cloud frame searching and matching method such as ICP (ITERATIVE CLOSEST POINT, iterative closest point algorithm) and the like has larger calculation amount because the data volume of the high-line number Lidar point cloud frame is huge and is of discrete data type, so that the practicability of the traditional method is greatly reduced when the number of the point cloud frames of the point cloud frame sample set is huge, and the real-time performance is poorer particularly on a low-calculation-force mobile platform. Disclosure of Invention In view of the above problems, a point cloud frame search matching method, apparatus, electronic device, and storage medium are proposed that overcome or at least partially solve the above problems, including: a point cloud frame search matching method, the method comprising: determining first description sub-information of each point Yun Zhen in the point cloud frame sample set, and determining first hash coding information corresponding to each first description sub-information; determining second description sub-information of an input point cloud frame, and determining second hash coding information corresponding to the second description sub-information; according to the first hash coding information and the second hash coding information, determining a similar candidate subset from the point cloud frame sample set, wherein the similar candidate subset comprises M point cloud frames, and M is a positive integer; and determining a target point cloud frame from the similar candidate subset according to the input point cloud frame. In some embodiments, the determining the first descriptor information of each point Yun Zhen in the point cloud frame sample set includes: Performing two-dimensional projection dimension reduction processing on the point cloud frame aiming at any point cloud frame in the point cloud frame sample set to generate a first two-dimensional matrix; And generating first descriptor information of the point cloud frame according to the first two-dimensional matrix. In some embodiments, the determining the first hash code information corresponding to each first descriptor information includes: And aiming at any point cloud frame in the point cloud frame sample set, carrying out hash coding on a first two-dimensional matrix in first description sub-information of the point cloud frame to obtain first hash coding information of the point cloud frame. In some embodiments, the performing hash encoding on the two-dimensional matrix in the first descriptor information of the point cloud frame to obtain first hash encoded information of the point cloud frame includes: Performing difference value downsampling on a first two-dimensional matrix in first description sub-information of the point cloud frame to obtain a second two-dimensional matrix; And performing discrete cosine transform on the second two-dimensional matrix to determine first hash coding information of the point cloud frame. In some embodiments, the determining a target point cloud frame from the similar candidate subset from the input point cloud frame comprises: calculating the similarity between the input point cloud frame and each point cloud frame in the similar candidate subset; And taking the point cloud frame with the largest similarity in the similar candidate subset as the target point cloud frame. In some embodiments, the calculating the similarity between the input point cloud frame and each point cloud frame in the similar candidate subset includes: and calculating the similarity between the second descriptor information and the first descriptor information of each point Yun Zhen in the similar candidate subset. In some embodiments, the calculating the similarity of the second descriptor information to the first descriptor information of each point Yun Zhen in the similar candidate subset includes: determining a second phase map of the second descriptor information and a first phase map of the first descriptor information of each point Yun Zhen in the similar candidate subset; and determining the similarity according to the first phase map and the second phase map. The embodiment of the application also provides a device for searching and matching the point cloud frame, which comprises the following steps: The first determining module is used for determining first description sub-information of each point Yun Zhen in the point cloud frame sample set and determining first hash coding information corresponding to each first desc