A Method for Automatical Extraction of Typical Disaster-bearing Targets from LiDAR Point Cloud in Coastal Zone

摘要

This paper presents a method based on georeferenced feature image to automatically extract typical disaster-bearing targets from coastal LiDAR data. Firstly, the noise and the water surface of LiDAR point cloud are removed by using the elevation histogram. Secondly, by analyzing the spatial distribution of point cloud, the georeferenced feature image of point cloud is generated. Finally, the image processing method and the corresponding relationship between the three-dimensional point cloud and the two-dimensional feature image are used to realize the automatic extraction of the targets. In this paper, the LiDAR cloud data of Haidian Island acquired by ALS70, is used as experimental data to verify the feasibility and practicability of the proposed method.