| The airborne LIDAR system is affected by the physical characteristics of the scanning equipment,scanning environment,system errors and integration errors,etc.,and the resulting original point cloud is often contaminated by noise.The massive amount of original point cloud data is directly used for subsequent algorithms.The efficiency is low.In addition,these noise points may also lead to calculation deviations.Therefore,it is necessary to filter,denoise,remove redundant points and streamline processing of massive point cloud data,so as to achieve rapid and efficient acquisition of valuable information from them In order to improve the processing efficiency and effect of subsequent computer algorithms,reduce memory requirements,etc.Filtering is the first step of point cloud processing,which is very important for subsequent processing.In addition,because the installation angle error has a great influence on the final mapping accuracy,it is necessary to check the installation angle error of the LIDAR system.Finally,because of the low flying height of the small UAV and the fact that the scanning angle of the laser cannot be set too large,the width of the terrain point cloud image collected for each continuous flight in a single channel is small,so it is necessary to use a reasonable and efficient point cloud stitching algorithm to Two point clouds with overlapping areas are spliced into a larger point cloud.In summary,the point cloud processing algorithm of the airborne LIDAR system studied in this paper is of great practical significance for the UAV LIDAR system independently developed by the laboratory to scan and obtain high-precision terrain point cloud maps.This paper designs and implements a set of high-precision and high-reliability airborne LIDAR scanning system,and studies the LIDAR point cloud processing algorithm,so as to realize various processing functions such as filtering,splicing and calibration of LIDAR point cloud The LIDAR system has the ability to complete the corresponding tasks.The main research contents of this article are as follows:(1)According to the needs of the project,complete the selection and debugging of each hardware sensor and the joint debugging based on the Raspberry Pi CM3,build a laser LIDAR hardware system that integrates each sensor;and complete the design of the LIDAR system software program,including embedded Linux acquisition program,ROS point cloud recovery procedures and ground station post-processing procedures,etc.(2)Analyze the error of the airborne LIDAR system and check the system installation angle and eccentricity error based on the point error model of the same name in the opposite flight zone and use linear interpolation to check the system synchronization error,etc..(3)A new point cloud smoothing smoothing method based on normal vector is proposed,which can perform smooth smoothing processing on the point cloud;and studies and implements curvature filtering,bilateral filtering and other commonly used filtering algorithms,etc.The point cloud data is processed by filtering,denoising optimization,removing redundant points and streamlining.(4)Research and implement the LIDAR point cloud stitching algorithm,and improve the classic ICP algorithm according to the actual situation of the system to greatly improve the efficiency and effect of the algorithm.(5)Based on QT + Open GL,a set of ground station software for airborne LIDAR system is designed and implemented to perform various processing operations on LIDAR point clouds;and design experiments to obtain experimental results through field flight data collection to verify the above point cloud processing The effectiveness of the algorithm.The airborne LIDAR system designed in this paper has been applied in actual projects,and its measurement accuracy meets the requirements of the project,and the error is within 10 cm. |
PDF Full Text Request