A time efficient algorithm for ground point filtering from mobile LiDAR data

Ground point filtering is a necessary step in case of accurate Digital Elevation Model (DEM) generation using Lidar data. In case of various feature extraction (including road objects also) from mobile Lidar data ground point filtering typically act as a preliminary step which might be a time consuming process. In this research paper we proposed a time efficient algorithm for ground point filtering from mobile Lidar data. The algorithm take input of raw Lidar data point cloud have only X, Y and Z values and perform the two dimensional projection and square X-Y gridding, algorithm also use the concept of height slicing of each grid to refine the rough ground points of each grid and leave the points have a Z value higher than a particular value and at last the algorithm use a concept of thresholded circular growing with thresholding parameter includes standard deviation of Z values to check whether a part of an object is of ground object or not. The algorithm is tested at one dataset and the corresponding ground points are detected having completeness of 99.9%.