MULTIGRID CELL SHAPE EVALUATION IN DTM FILTERING OF MOBILE LIDAR POINT CLOUDS

Abstract. The georeferenced information acquisition for large scales digital terrain models generation is usually a rather time-consuming and costly process. The use of point clouds gathered by mobile LiDAR systems therefore appears as a possible source for the creation of those models, particularly in remote areas where security issues requires a maximum speed field collection information. However, the filtering challenge to separate the points that represents the terrain surface from the remaining, in an automatic and consistent way, is an open challenge that continues to arouse the interest of researchers. Using a multigrid filtering method, this work presents a comparative analysis of different shapes of regular grid cells, including the implementation algorithms for each cell shape. Also, a coordinate grid concept applied to each cell shape is proposed. Finally, a discussion about the obtained results of the method application using point clouds collected in different environments from different ground LiDAR systems are presented.