Research On LiDAR Points Filtering Algorithm Based On Morphological Triangulated Irregular Network

LiDAR is a new kind of measurement technology which can quickly obtain the spatial geographic information.Comparing with the traditional methods of measurement and remote sensing,it has the unique superiority.LiDAR technology has developed rapidly in recent years,it has been used to different industries in the field of surveying and mapping work,and has been widely concerned.Point cloud filtering technology is the basic work in LiDAR data processing,it is a key role in DEM production as a frontier subject for technology development and research.This paper studys a methods around the airborne LiDAR point cloud data filtering.The main research and innovation are as follows.Firstly,aiming at the influence of the coarse difference point on the filtering result of the airborne LiDAR point cloud data,this paper presents a method of gross error elimination which based on KD tree tissue cloud data,and the result of validity which use some selected test data provided by ISPRS website to verify the method of gross error elimination was provided the guarantee for the follow-up point cloud filter.Secondly,it was a difficult problem to select more the initial seed point of traditional triangulation,we propose a method of filtering algorithm combining morphological and irregular triangles from "coarse" to "fine".We analyze the influence of different structural element windows to the morphological filtering results,and use the morphological coarse filter as the basic data of next irregular triangulation filter.Lastly,improving the process of iterative network of the irregular triangulation filter method.We determine seed points on the basis of the coarse filter.The improved irregular triangulation filter used to fine filter the point cloud data and obtain the ground point,reduced the number of network construction and error accumulation.By comparing the result of the algorithm with morphological and irregular triangulation use filtering test data provided by ISPRS,can find that the algorithm can obtain more accurate filtering results and the DEM generated by filtering result is more close to the real surface.