Application of Airborne Radar Classification Mapping Method in Mudflat Mapping

This paper aims to study how to efficiently achieve large-area lidar mapping of intertidal zones and more accurately reflect their topographic structure. It employs a terrain module classification and modeling method to classify and reorganize terrain structures such as woodland, plains, and hills. The BP neural network toolbox in MATLAB software is used to call mathematical models to train and learn the above various terrain modules, which are then applied to the overall point cloud data modeling analysis of Dongshan Island. Through field research and comparative analysis with remote sensing mini-plan views, it is concluded that using the terrain module classification and modeling method for laser point cloud modeling can better reflect the structural changes and topographic relief characteristics of complex intertidal zone terrains.