Dense Point Cloud Classification in Terrain Modeling

Abstract

The article considers the existing algorithms for classifying elements of a surface model. The procedure for photogrammetric processing of images of unmanned aerial vehicles is given. The equipment necessary for the preparation of initial materials of aerial photography for photogrammetry is described. The features of processing data obtained from unmanned aerial vehicles using the «Agisoft Metashape» software package are presented. A new approach to the classification of elements of a surface model and grouping similar elements into separate layers of a dense point cloud is proposed. The possibility of constructing a reliable digital elevation model based on aerial photography in a fully automatic mode is analyzed. The results of the work carried out on the connection «geodesy-general plan-building» were summed up and the positive and problematic aspects of the use of new information technologies in design.