A filtering method for generating DTM based on multi-scale mathematic morphology

A filtering method is proposed in this paper that can generate DTM (digital terrain model) based on multi-scale mathematic morphology. With the second echo information of LIDAR, the method firstly extracts the edge point of vegetation and house, and then process opening operation of mathematical morphology with the remainder data in local region. Comparing adjacent filter surface with the threshold parameter, we extract surface feature during gradually increasing transform scale of filtering window. Experiment shows that our method can effectively generate DTM and also can maintain the terrain details.