CN-122023134-A - Filtering processing method, device and equipment for tailing point cloud and readable storage medium

CN122023134ACN 122023134 ACN122023134 ACN 122023134ACN-122023134-A

Abstract

The invention provides a filtering processing method, device and equipment for a tailing point cloud and a readable storage medium, which comprise the steps of obtaining the laser point cloud generated in the working process of a rotary mirror type laser radar, detecting edge areas of two detection objects through a sliding filtering window, judging whether the tailing point cloud exists in the sliding filtering window, dividing the tailing point cloud based on a distance threshold if the tailing point cloud exists, generating the tailing point cloud to be processed with a near-view target distance and a far-view target distance, judging whether the tailing point cloud to be processed is deleted directly, if yes, deleting the tailing point cloud to be processed directly, and if not, executing the filtering processing of the tailing point cloud to be processed and outputting the point cloud after the filtering processing. According to the invention, the object contour edge is extracted through the sliding filter window, so that the tail point cloud identification accuracy and processing efficiency are improved, meanwhile, the point cloud information of the object edge is reserved through restoring the tail point cloud, and the point cloud quality is improved.

Inventors

ZHANG MINGYONG
JIA ZHIJIE
Bai Zhongshan
YANG KAIHONG

Assignees

成都晶品夜视光电科技有限公司

Dates

Publication Date: 20260512
Application Date: 20251226

Claims (10)

1. The filtering processing method of the tailing point cloud is characterized by comprising the following steps of: Acquiring laser point clouds generated in the working process of the rotating mirror type laser radar; detecting edge areas of two detection objects through a sliding filter window, and judging whether trailing point clouds exist in the sliding filter window; If a tailing point cloud exists, dividing the tailing point cloud based on a distance threshold value to generate a to-be-processed tailing point cloud with a near target distance and a far target distance; Judging whether the tailing point cloud to be processed is directly deleted, if so, directly deleting the tailing point cloud to be processed, and if not, executing the filtering processing of the reduction to the tailing point cloud to be processed and then outputting the point cloud after the filtering processing.
2. The method for filtering the tailing point cloud according to claim 1, wherein before the step of detecting the edge areas of the two detection objects by the laser point cloud through the sliding filter window, the sliding filter window is established according to the type of a coordinate system in which the laser point cloud is located, if the sliding filter window is a polar coordinate system, the sliding filter window is arranged along the horizontal corner and the pitching corner of the laser point cloud, and if the sliding filter window is a rectangular coordinate system, the sliding filter window is arranged along the horizontal coordinate direction and the vertical coordinate direction.
3. The method for filtering the tailing point cloud according to claim 1, wherein the step of detecting edge areas of two detection objects by sliding the filtering window, and determining whether the tailing point cloud exists in the sliding filtering window specifically includes: Setting window size and moving step length parameters of the sliding filter window, performing edge detection, and calculating absolute values of distance differences of adjacent points in window areas along the horizontal direction, the vertical direction or +/-45 DEG direction of rotating mirror scanning; Acquiring the number of the point clouds with the absolute value in a preset interval, and judging that trailing point clouds exist in the window area if the number of the point clouds is larger than a set threshold value; The preset interval is a range interval in which the distance between two detection objects is located, and the threshold value is a proportional value of the number of point clouds in the window area.
4. The method for filtering a tailing point cloud according to claim 3, wherein if a tailing point cloud exists, dividing the tailing point cloud based on a distance threshold, and generating a tailing point cloud to be processed having a near target distance and a far target distance specifically includes determining a midpoint target distance r 0 in the window area, and dividing the tailing point cloud in the window area into points greater than the midpoint target distance and less than or equal to the midpoint target distance; The midpoint target distance r 0 is a distance intermediate value or average value of the trailing point cloud in the window area, a point smaller than or equal to the midpoint target distance is taken as a close-range target distance r 1 , and a point larger than the midpoint target distance is taken as a far-range target distance r 2 ,r 1< r 0< r 2 .
5. The method for filtering a tailing point cloud according to claim 4, wherein the step of determining whether to delete the tailing point cloud to be processed directly, if yes, deleting the tailing point cloud to be processed directly, and if no, outputting the filtered point cloud after the filtering process of performing the restoration on the tailing point cloud to be processed specifically includes: If the tailing point cloud to be processed is deleted directly, deleting the point between the near target distance r 1 and the distant target distance r 2 directly; If not, executing filtering processing for restoring the tail point cloud to be processed, and calculating a restored distance r j_new according to a distance r j of any point between the near target distance r 1 and the far target distance r 2 : if r j >r 0 , the distance between any point after the reduction is r j_new = r j ×A +r 2 XB If r j <r 0 , the distance between any point after the reduction is r j_new = r j ×A +r 1 XB Wherein A and B are proportionality coefficients, , ,A+B=1。
6. The method according to claim 1, further comprising moving the sliding filter window to a next position after the step of performing the filtering process of deleting or restoring the to-be-processed tailing point cloud, and continuing to perform the filtering process of the tailing point cloud until all laser point cloud processing is completed.
7. The filtering processing method of the tailing point cloud as claimed in claim 1, wherein the size of the sliding filter window is set to 7 x 7 or 11 x 11.
8. A filtering apparatus for a tailing point cloud, configured to implement the filtering method for a tailing point cloud according to any one of claims 1 to 7, comprising: the acquisition module is used for acquiring laser point clouds generated in the working process of the rotating mirror type laser radar; The judging module is used for detecting the edge areas of the two detection objects through the sliding filter window and judging whether trailing point clouds exist in the sliding filter window or not; The segmentation module is used for segmenting the tailing point cloud based on a distance threshold value if the tailing point cloud exists, and generating the tailing point cloud to be processed with a near target distance and a distant target distance; And the processing module is used for judging whether the tailing point cloud to be processed is directly deleted, if yes, the tailing point cloud to be processed is directly deleted, and if not, the tailing point cloud to be processed is subjected to the restored filtering processing and then the point cloud after the filtering processing is output.
9. An electronic device comprising a memory and a processor, the memory having stored thereon a computer program, characterized in that the processor, when executing the computer program, implements the method according to any of claims 1-7.
10. A computer readable storage medium, on which a computer program is stored, which computer program, when being executed by a processor, implements the method according to any of claims 1-7.

Description

Filtering processing method, device and equipment for tailing point cloud and readable storage medium Technical Field The present invention relates to the field of data processing, and in particular, to a method, an apparatus, a device, and a readable storage medium for filtering a tailing point cloud. Background The laser radar is photoelectric equipment capable of measuring three-dimensional scene information of surrounding environment, and is widely applied to the fields of intelligent driving, topographic mapping, robot navigation, obstacle avoidance and the like. The working principle is that one or more laser beams are emitted, and the space distance information of the object is obtained after the laser signals reflected by the object are received. These spatial distance information are presented in the form of point cloud data, each point cloud storing data in the form of a cartesian coordinate system or a spherical polar coordinate system. Because the laser is emitted as a beam and has a certain divergence angle, when a single laser beam is detected to reach two objects near one far away, tailing point clouds easily appear in laser point cloud data, so that a ranging error is caused. In the prior art, as the application number 202311311755.9, the patent name is the processing method, the device, the equipment and the readable storage medium of the point cloud tailing point, and the application number 202111040306.6, the patent name is the removing method, the device, the electronic equipment and the readable storage medium of the point cloud tailing point all provide corresponding solutions for the processing of the tailing point cloud, however, in the existing solutions, most of the field detection methods based on the distance threshold and the angle threshold or the outlier detection algorithm are adopted, some solutions are added with the light energy intensity value to judge the tailing point, and the tailing point is removed by setting a reasonable threshold range, and usually only the point cloud data judged as the tailing point is removed. The technical measures have simple criteria, but poor robustness, can have the problems of deleting some effective points by mistake or deleting inexhaustible points, and the schemes only remove tailing points, can not restore the edges of objects, and can cause the inherent defects of losing point cloud of the edges of the objects and reducing point cloud data. Disclosure of Invention In order to solve the technical problems, the invention provides a method, a device, equipment and a readable storage medium for filtering a tailing point cloud, which are used for judging the tailing point cloud by detecting the edge profile of an object based on edge filtering of three-dimensional data and deleting or restoring the tailing point cloud, so that the edge point cloud data of a detected object are reserved as completely as possible, and the quality of the point cloud is improved. In a first aspect, the present invention provides a filtering processing method for a tailing point cloud, which specifically includes the following steps: Acquiring laser point clouds generated in the working process of the rotating mirror type laser radar; detecting edge areas of two detection objects through a sliding filter window, and judging whether trailing point clouds exist in the sliding filter window; If a tailing point cloud exists, dividing the tailing point cloud based on a distance threshold value to generate a to-be-processed tailing point cloud with a near target distance and a far target distance; Judging whether the tailing point cloud to be processed is directly deleted, if so, directly deleting the tailing point cloud to be processed, and if not, executing the filtering processing of the reduction to the tailing point cloud to be processed and then outputting the point cloud after the filtering processing. In one embodiment, before the step of detecting the edge areas of the two detection objects by the laser point cloud through the sliding filter window, the method further comprises the steps of establishing the sliding filter window according to the type of a coordinate system where the laser point cloud is located, setting the sliding filter window along the horizontal corner and the pitching corner direction of the laser point cloud if the laser point cloud is in a polar coordinate system, and setting the sliding filter window along the horizontal coordinate direction and the vertical coordinate direction if the laser point cloud is in a rectangular coordinate system. In one embodiment, the step of detecting edge areas of two probes by the sliding filter window to determine whether a tailing point cloud exists in the sliding filter window specifically includes: Setting window size and moving step length parameters of the sliding filter window, performing edge detection, and calculating absolute values of distance differences of adjacent points in window areas