CN-116953658-B - Data processing method, device and storage medium

CN116953658BCN 116953658 BCN116953658 BCN 116953658BCN-116953658-B

Abstract

The invention discloses a data processing method, a device and a storage medium, wherein the method comprises the steps of collecting a frame of laser point cloud data by using a laser radar, and correcting the laser point cloud data to obtain corrected laser point cloud data; the method comprises the steps of carrying out straight line detection on laser point clouds included in corrected laser point cloud data to obtain straight lines corresponding to each laser point cloud included in the corrected laser point cloud data, setting corresponding weights according to distances between the laser point clouds included in the corrected laser point cloud data and the corresponding straight lines, and determining positioning information and actual laser point cloud data of a laser radar based on the corrected laser point cloud data and the weights of each laser point cloud included in the corrected laser point cloud data by utilizing a point-to-line nearest iteration algorithm. By the technical scheme, the accuracy of data processing is improved.

Inventors

Wan Yaozhong
LIAO WENWEI
YANG ZEXIAN

Assignees

中移（杭州）信息技术有限公司
中国移动通信集团有限公司

Dates

Publication Date: 20260505
Application Date: 20221021

Claims (12)

1. A method of data processing, the method comprising: collecting a frame of laser point cloud data by using a laser radar, and correcting the laser point cloud data to obtain corrected laser point cloud data; Performing straight line detection on laser point clouds included in the corrected laser point cloud data to obtain straight lines corresponding to each laser point cloud included in the corrected laser point cloud data; Setting a corresponding weight according to the distance from a corresponding straight line for each laser point cloud included in the corrected laser point cloud data; And determining positioning information of the laser radar and actual laser point cloud data based on the corrected laser point cloud data and the weight of each laser point cloud included in the corrected laser point cloud data by using a point-to-line nearest iterative algorithm.
2. The method of claim 1, wherein the modifying the laser point cloud data to obtain modified laser point cloud data comprises: Acquiring pose change information of the laser radar in the process of acquiring the laser point cloud data by the laser radar; And correcting the laser point cloud data based on the pose change information to obtain corrected laser point cloud data.
3. The method according to claim 2, wherein the step of acquiring pose change information of the laser radar during the step of acquiring the laser point cloud data by the laser radar includes: Acquiring inertial measurement unit data and/or laser radar odometer data; And determining the pose change information according to the inertial measurement unit data and/or the laser radar odometer data by using a nonlinear Kalman filtering method.
4. The method of claim 2, wherein the pose change information comprises initial pose information when a first laser point cloud is acquired and final pose information when a last laser point cloud is acquired in the process of acquiring the laser point cloud data by the laser radar, and the correcting the laser point cloud data based on the pose change information to obtain corrected laser point cloud data comprises: Dividing the acquisition time length of the laser radar for acquiring the laser point cloud data into a plurality of sub-time lengths; for each duration in the plurality of sub-durations, estimating corresponding pose information by adopting a linear interpolation mode according to the initial pose information and the final pose information; acquiring laser point clouds acquired in each time length in the plurality of sub-time lengths from the laser point cloud data to obtain a plurality of groups of laser point clouds corresponding to the plurality of sub-time lengths one by one; Determining pose information corresponding to the same sub-time length as pose information corresponding to each laser point cloud in the group aiming at each laser point cloud in the plurality of groups of laser point clouds; And aiming at each laser point cloud included in the laser point cloud data, taking the coordinate system of the first laser point cloud as a reference coordinate system, and determining the corresponding laser point cloud data under the reference coordinate system by utilizing the corresponding pose information to obtain the corrected laser point cloud data.
5. The method according to claim 1, wherein the performing straight line detection on the laser point clouds included in the corrected laser point cloud data to obtain straight lines corresponding to each laser point cloud included in the corrected laser point cloud data includes: selecting a first laser point cloud from laser point clouds included in the corrected laser point cloud data, and determining the first laser point cloud as a1 st laser point cloud; Sequentially selecting laser point clouds from the corrected laser point cloud data according to a preset direction by taking the 1 st laser point cloud as a reference until the kth laser point cloud is selected, wherein a straight line residual error determined based on the 1 st laser point cloud to the kth laser point cloud is greater than a preset residual error threshold value, and k is a natural number greater than 1; Determining a straight line determined based on the 1 st to the k-1 st laser point clouds as a straight line corresponding to each of the 1 st to the k-1 st laser point clouds; And continuously detecting the laser point clouds included in the corrected laser point cloud data according to the preset direction by taking the kth laser point cloud as a reference until a straight line corresponding to each laser point cloud included in the corrected laser point cloud data is obtained.
6. The method according to claim 5, wherein the sequentially selecting laser point clouds from the corrected laser point cloud data in a preset direction based on the 1 st laser point cloud until the kth laser point cloud is selected, and the straight line residual determined based on the 1 st laser point cloud to the kth laser point cloud is greater than a preset residual threshold value, comprises: Sequentially selecting laser point clouds from the corrected laser point cloud data according to the preset direction by taking the 1 st laser point cloud as a reference until the nth laser point cloud is selected, wherein the linear parameter determined based on the 1 st laser point cloud to the n-1 st laser point cloud and the linear parameter determined based on the 1 st laser point cloud to the nth laser point cloud are greater than a preset parameter threshold, and n is a natural number greater than 1 and less than or equal to k; removing the nth laser point cloud from the corrected laser point cloud data, and re-determining the next selected laser point cloud as the nth laser point cloud; and sequentially selecting laser point clouds from the corrected laser point cloud data according to the preset direction until the kth laser point cloud is selected, wherein the linear residual error determined based on the 1 st laser point cloud to the kth laser point cloud is larger than the preset residual error threshold value.
7. The method of claim 5, wherein the continuing to perform straight line detection on the laser point cloud included in the corrected laser point cloud data according to the preset direction based on the kth laser point cloud includes: And determining a straight line determined based on the m-th laser point cloud to the last laser point cloud as a straight line corresponding to each of the m-th laser point cloud to the last laser point cloud under the condition that the straight line residual determined based on the m-th laser point cloud to the last laser point cloud is smaller than or equal to the preset residual threshold, and the straight line residual determined based on the m-th laser point cloud to the last laser point cloud is larger than the preset residual threshold, wherein m is a natural number larger than k.
8. The method of claim 5, wherein the continuing to perform straight line detection on the laser point cloud included in the corrected laser point cloud data according to the preset direction based on the kth laser point cloud includes: and determining a straight line determined based on the 1 st to the k th laser point clouds and the m to the last laser point clouds as a straight line corresponding to each of the 1 st to the k th laser point clouds and the m to the last laser point clouds, wherein m is a natural number greater than k, when a straight line residual determined based on the selected m to the last laser point clouds is less than or equal to the preset residual threshold and a straight line residual determined based on the 1 st to the k th to the last laser point clouds is less than or equal to the preset residual threshold.
9. The method of claim 5, wherein the continuing to perform straight line detection on the laser point cloud included in the corrected laser point cloud data according to the preset direction based on the kth laser point cloud includes: And determining a straight line determined based on the m-th laser point cloud to the last laser point cloud as a straight line corresponding to each of the m-th laser point cloud to the last laser point cloud under the condition that the straight line residual determined based on the m-th laser point cloud to the last laser point cloud selected is smaller than or equal to the preset residual threshold, and the straight line residual determined based on the 1-th laser point cloud to the k-th laser point cloud and the m-th laser point cloud is larger than the preset residual threshold, wherein m is a natural number larger than k.
10. A data processing apparatus, comprising: The acquisition module is used for acquiring a frame of laser point cloud data by using the laser radar and correcting the laser point cloud data to obtain corrected laser point cloud data; The detection module is used for carrying out straight line detection on the laser point clouds included in the corrected laser point cloud data to obtain straight lines corresponding to each laser point cloud included in the corrected laser point cloud data; The setting module is used for setting corresponding weights according to the distance between each laser point cloud included in the corrected laser point cloud data and the corresponding straight line; the determining module is used for determining positioning information of the laser radar and actual laser point cloud data based on the corrected laser point cloud data and weight of each laser point cloud included in the corrected laser point cloud data by using a point-to-line nearest iterative algorithm.
11. A data processing device is characterized by comprising a processor, a memory and a communication bus; The communication bus is used for realizing communication connection between the processor and the memory; The processor being adapted to execute a computer program stored in the memory for implementing the data processing method of any of claims 1-9.
12. A computer-readable storage medium, wherein the computer-readable storage medium stores one or more computer programs, the one or more computer programs may be executed by one or more processors to implement the data processing method of any of claims 1-9.

Description

Data processing method, device and storage medium Technical Field The present application relates to the field of robotics, and in particular, to a data processing method, apparatus, and storage medium. Background With the rapid development of the robot industry in recent years, the requirements of various industries on mobile robots are becoming wider and wider. Therefore, mobile robots need to be deployed into a variety of industrial applications, while mobile robots that can accommodate a variety of changing environments, are low cost and that can function properly therein would be favored. The two-dimensional laser radar is used as environment sensing equipment with low cost and is widely applied to the mobile robot. However, as the frequency of the two-dimensional laser radar for collecting the point cloud data is lower, the mobile robot has shaking conditions in the moving process, or as the current scene has long distance or small volume objects, the collected point cloud data contains points with offset errors, the accuracy is lower, and the point cloud data with lower accuracy can influence the drawing precision and the positioning accuracy of the mobile robot. Disclosure of Invention In order to solve the above technical problems, embodiments of the present invention are expected to provide a data processing method, apparatus, and storage medium, which improve accuracy of data processing. The technical scheme of the invention is realized as follows: the invention provides a data processing method, which comprises the following steps: collecting a frame of laser point cloud data by using a laser radar, and correcting the laser point cloud data to obtain corrected laser point cloud data; Performing straight line detection on laser point clouds included in the corrected laser point cloud data to obtain straight lines corresponding to each laser point cloud included in the corrected laser point cloud data; Setting a corresponding weight according to the distance from a corresponding straight line for each laser point cloud included in the corrected laser point cloud data; And determining positioning information of the laser radar and actual laser point cloud data based on the corrected laser point cloud data and the weight of each laser point cloud included in the corrected laser point cloud data by using a point-to-line nearest iterative algorithm. In the above method, the correcting the laser point cloud data to obtain corrected laser point cloud data includes: Acquiring pose change information of the laser radar in the process of acquiring the laser point cloud data by the laser radar; And correcting the laser point cloud data based on the pose change information to obtain corrected laser point cloud data. In the above method, in the process of acquiring the laser point cloud data by the lidar, pose change information of the lidar includes: Acquiring inertial measurement unit data and/or laser radar odometer data; And determining the pose change information according to the inertial measurement unit data and/or the laser radar odometer data by using a nonlinear Kalman filtering method. In the method, the pose change information comprises initial pose information when a first laser point cloud is acquired and final pose information when a last laser point cloud is acquired in the process of acquiring the laser point cloud data by the laser radar, and the correcting the laser point cloud data based on the pose change information to obtain corrected laser point cloud data comprises the following steps: Dividing the acquisition time length of the laser radar for acquiring the laser point cloud data into a plurality of sub-time lengths; for each duration in the plurality of sub-durations, estimating corresponding pose information by adopting a linear interpolation mode according to the initial pose information and the final pose information; acquiring laser point clouds acquired in each time length in the plurality of sub-time lengths from the laser point cloud data to obtain a plurality of groups of laser point clouds corresponding to the plurality of sub-time lengths one by one; Determining pose information corresponding to the same sub-time length as pose information corresponding to each laser point cloud in the group aiming at each laser point cloud in the plurality of groups of laser point clouds; And aiming at each laser point cloud included in the laser point cloud data, taking the coordinate system of the first laser point cloud as a reference coordinate system, and determining the corresponding laser point cloud data under the reference coordinate system by utilizing the corresponding pose information to obtain the corrected laser point cloud data. In the above method, the detecting the straight line of the laser point cloud included in the corrected laser point cloud data to obtain a straight line corresponding to each laser point cloud included in the corrected laser point cloud data include