Review of Automatic Processing of Topography and Surface Feature Identification LiDAR Data Using Machine Learning Techniques

Machine Learning (ML) applications on Light Detection And Ranging (LiDAR) data have provided promising results and thus this topic has been widely addressed in the literature during last few years. This paper reviews essential more recent completed studies topography surface feature identification domain. Four areas, with respect to suggested approaches, analyzed discussed: input data, concepts of point cloud structure for applying ML, ML techniques used, LiDAR data. Then, an overview is underline advantages disadvantages research axis. Despite training labelling problem, calculation cost, undesirable shortcutting due downsampling, most proposed methods use supervised classify downsampled Furthermore, despite occasional highly accurate results, cases still require filtering. In fact, a considerable number adopted approaches same employed image processing profit from available informatics tools. Knowing that clouds represent rich 3D effort needed develop specialized