Study On Airborne LiDAR Point Cloud Data Post-Processing Theory And Technology

As a new active remote sensing technology, airborne LiDAR has produced the gross breakthrough in the aspect of three-dimensional earth spatial information acquisition, and has caused great interests in the surveying and mapping, forestry and other related industries. In recent years, with the development of related technologies and the expansion of the social needs, airborne LiDAR which represents a new development direction in the field of earth observation, offers a kind of completely new technical means for the high temporal-spatial resolution earth spatial information acquisition, make data acquisition, processing automation and intellectualization to be possible. Compared with the rapid development of hardware technology, the post-processing algorithms (such as automatic filtering and classification) of LiDAR point cloud data are still in the stage of being researched, and need to be further studied.This paper systematically discusses the airborne LiDAR system and the theories and techniques of data post-processing. The major innovations are listed below:(1) Data management. Large-scale point cloud data can be fast accessed by establishing regular grid index,and can also be read and written rapidly on the basis of memory-mapped file technology, drew quickly based on the "mesh" geometry which used to simulate dense point cloud data,and can be real-time refreshed with the method of block mapping and sparse displaying.(2) Filtering algorithm. Mathematical morphological and progressive TIN filtering algorithms have been improved in this paper. Experiments proves that the modified algorithms are stable and efficient with good reliability and practicability.(3) Point cloud data post-processing technology. The TopLiDAR that used to process LiDAR point cloud data,which is designed and developed based on MicroStation. It has mathematical morphological and progressive TIN filtering algorithms which have been improved, and has manual classification tools included point,line, surface, continuous dynamic profile editing, etc., which improve the efficiency of LiDAR point cloud data filtering and classification.Studying filtering and classification algorithms and developing the appropriate post-processing software of the LiDAR point cloud data are one of the effective ways to improve its data accuracy, also one of the researched focuses in the field of its data post-processing. The research in this paper may be helpful for it.