Ground and Non-Ground Filtering for Airborne LIDAR Data

Abstract

Automatic ground filtering for Light Detection and Ranging (LIDAR) data is a critical process for Digital Terrain Model and three-dimensional urban model generation. Various methods have been proposed in literature to separate ground from non-ground, but sometimes problem has been occurred due to the similar characteristics possessed by ground and non-ground objects. The proposed approach in this paper is based on neighborhood based approach. Hierarchy of preprocessing is done for LIDAR data using various essential tools. K-D tree is used to distinguish the bare ground and non-ground objects using nearest neighbor search. Experimental results show the effectiveness of the proposed approach.