A Robust Filtering Algorithm of LiDAR Data for DTM Extraction

摘要

For DTM extraction of road linear project from LiDAR data, this paper proposes a robust ground filtering algorithm. Firstly, a pyramidal hierarchical structure of ground candidate points is formed. Secondly, in each level, the Thin Plate Spline (TPS) surface interpolation is iteratively used to produces smooth surfaces, and based on the point's residuals from the surface the ground candidate points are inspected. The experiments show that this method correctly determines DTM even in those cases of heavy vegetation, and achieve a better result for the road.