Research On Airborne LiDAR Point Clouds Devegetation Algorithm In Forest Area Under Complex Terrain

Airborne laser radar(LiDAR)technology has been widely used in many fields such as digital ground model(DEM),infrastructure construction,and forest parameter extraction,especially in the investigation and utilization of forest resources in mountainous forest areas with complex terrain.important role.Compared with traditional ground technology,LiDAR technology has the characteristics of all-weather and high operating efficiency,but dense vegetation occlusion and various non-ground points also seriously affect the ability of airborne radar to obtain real ground points.Therefore,effective devegetation filtering of airborne LiDAR point cloud data has become a key technology that restricts the application of airborne LiDAR technology in complex mountain forest terrain.In this paper,the point cloud data of the research area is firstly processed into a grid,and on this basis,terrain factors are introduced,and two improved single filtering algorithms are proposed,namely,the adaptive slope filtering algorithm and the improved cloth simulation filtering algorithm,and further A combined filtering algorithm combining the above two improved algorithms is proposed.The new algorithm was studied and verified by using the power line in the forest area with complex mountainous terrain and the banded point cloud data on both sides in Yinjiang,Guizhou.The results show that the original single slope filtering algorithm and the cloth simulation filtering algorithm have significant Improve the effect,and the combined filtering effect is better than the two improved single filtering algorithms.The specific work and achievements are as follows:(1)The technical principles and characteristics of the airborne LiDAR system are systematically expounded,the research progress of the airborne LiDAR point cloud data filtering algorithm is sorted out,and the slope-based filtering algorithm and the principle of the cloth simulation algorithm(CSF)are analyzed,clarifying this Limitations of the two algorithms.Aiming at the problem that the slope filtering algorithm adopts a fixed slope threshold and the cloth simulation algorithm adopts fixed parameters for internal force constraints,it is difficult to apply to complex mixed terrain filtering.An improved method is proposed: firstly,the point cloud data is properly gridded to obtain the points of the research area.Cloud multi-scale virtual grid,on this basis,introduces terrain factors,and proposes to use terrain factors to adaptively judge and set the slope threshold and key parameters of cloth simulation filtering in the grid area of 8 domains,thus enhancing the performance of the two filtering algorithms.Adaptability to complex terrain areas.The point cloud data of Yinjiang power line strip area is filtered,and the results show that the filtering effect of the two improved algorithms,which use the terrain factor to adaptively set the slope threshold and the key parameters of the cloth simulation algorithm,is significantly improved.Fractured mountainous forest areas(especially dense areas with tall vegetation)have achieved relatively good filtering results.(2)In view of the fact that the adaptive slope filtering algorithm has poor filtering effect in low vegetation areas with broken terrain,but has strong filtering ability in steep areas with rapid slope changes,and the improved cloth simulation filtering algorithm with adaptive parameter settings can be used in most areas.It is good at filtering non-ground points,but it is easy to mistakenly remove ground points in the cliff and cliff areas with sudden changes in terrain,resulting in terrain distortion.A combined filtering algorithm that combines the above two improved single filtering algorithms is proposed.The specific implementation process of the combined filtering algorithm is as follows: firstly,the point cloud is subjected to preliminary grid processing,and the angle between the vector formed by the lowest point in the grid and the lowest points in the 8 neighborhoods and the X-Y plane,and the point cloud density in the grid In order to judge the topography in the self-adaptive judgment grid,the judgment result is used as the basis for whether to proceed to the next level of grid processing,and the above-mentioned combined filtering process is performed on the point cloud data of each level of grid.The results of this paper show that the combined filtering algorithm realizes the complementary advantages of the two single improved filtering algorithms,and the filtering results maintain the integrity of the terrain to the greatest extent.(3)The combined filtering algorithm proposed in this paper is used to generate the DEM of the Yinjiang power line banded mountainous forest area and the power line extraction experiment.The experimental results show that the high-precision DEM of the research area can be obtained by using the combined filtering algorithm proposed in this paper,and the extracted power line image is complete and clear.