ABSTRACT
In recent years, geospatial feature extraction has become a very widely recognized and actively pursued research area. Various geospatial feature extraction algorithm exsist which involve the filtering of ground points from Light Detection and Ranging (LiDAR) point cloud datasets as the basic and imperative step. In the case of the LiDAR dataset, the accuracy of the generated Digital Terrain Model (DTM) is dependent on the accuracy of the ground filtering method. The objective of this research paper is to propose a novel approach for automated ground point filtering and generation of Digital Terrain Model (DTM) using the LiDAR point cloud dataset. The Proposed approach involves five steps such as segmentation of the LiDAR point cloud, vertical slicing, slope based filtering, circular growing, and DTM generation. ArcGIS 10.3 has been used for the generation of DTM using filtered ground points and also a Least Significant Bit (LSB) based image steganography technique is applied to secure the DTM. The proposed method has been tested at two different datasets. Dataset 1 is captured using mobile LiDAR, while Dataset 2 is captured using terrestrial LiDAR. Ground points of both datasets are filtered out with an average value of completeness and correctness are 98.11%, 94.71% respectively and corresponding DTM is generated.
