Point Cloud Data Processing Method for Bulk Cargo Piles Based on Airborne LiDAR

Aiming at the problem of limited measurement range, unstable accuracy, and lack of flexibility in the current tally level of the bulk cargo terminal, a rapid and accurate processing method of bulk cargo pile point cloud data based on the application of airborne light detection and ranging (LiDAR) in bulk cargo terminal was proposed in this paper. Firstly, the collected dataset was preprocessed by this method, and an improved voxel grid adaptive downsampling algorithm was used, which can adaptively reduce the number of sampling points according to the density distribution of point cloud data. Secondly, for the semantic segmentation problem of point cloud data, a progressive two-stage processing method was used that combined the two algorithms to more accurately extract the point cloud data of the target pile. The experimental results show that: This method is not only applicable to the environment of the bulk cargo terminal, but also can accurately and quickly extract the point cloud data of the target bulk cargo pile.