A Spatial Clustering Filtering Method for Airborne LiDAR Point Cloud Based on Dual Distance

Ground filtering for airborne light detection and ranging (LiDAR) is one of the core steps to obtain digital terrain model (DTM) in complex terrain. In view of the deficiencies of excessive erosion as well as low automation in existing filtering method, this paper proposes a spatial clustering filtering (SCF) method. Firstly, the point cloud data is described as the Octree index structure, and then segmented based on plane fitting. After that, a coarse spatial clustering process based on dual distance is implemented to obtain a set of seed points. Secondly, a triangular irregular network (TIN) is built based on the classic progressive densification method with a set of seed points, and the rest of the valid point clouds are classified iteratively. Finally, the experimental results show that the SCF method in this paper can not only avoids the single threshold restrictions to the filtering result, but also maintains complicated topography when the ground information in the discontinuous terrain is filtered, all of the above are contribute to improve the reliability of the filtering result.