Research On The Ground Filtering Method Of Complex Terrain In 3D Laser Point Cloud

The 3D laser scanning technology has the advantages of high precision,high efficiency and non-contact measurement.It can quickly obtain the 3D coordinate information of the surface and the surface of the object,meet the rapid growth of 3D geographic information in various industries,and is widely used in the field of surveying and mapping.The point cloud data obtained by the 3D laser scanning system contains different terrain and rich feature information.The ground points are separated from the feature points by ground filtering,and high-precision terrain data can be obtained for research and application.The actual terrain is divided into flat,gentle and steep slopes according to the slope.For complex terrain data with different terrains or terrain fluctuations,the spatial distribution of terrain points is scattered,there are no obvious rules for elevation changes,and there are interference noise points such as ground objects,which makes it difficult to obtain high-precision filtering results.This article analyzes the spatial distribution characteristics of different terrain data in depth,uses slopes to identify changes in terrain data,and studies ground filtering for complex terrain data.The main contents and innovations are as follows:(1)To sort out the development and trend of 3D laser scanning system,summarize the advantages and disadvantages of existing ground filtering algorithms,introduce the characteristics of point cloud data,and lay a theoretical foundation for the research of filtering methods.(2)Aiming at the problem of irregular changes in complex terrain,the slope feature terrain segmentation method is studied.Calculate the slope value,use the slope value variance and average value to describe the terrain characteristics,divide the point cloud data into different categories,and filter them according to different terrain characteristics.(3)An adaptive slope ground filtering method is designed to perform ground filtering on gentle slope terrain.According to different terrain types,filter windows of corresponding scales are selected,and the statistical characteristics of slope values are used to automatically obtain slope threshold parameters and obtain complete ground point data.(4)Fitting and filtering the profile curve of point cloud data on steep slope terrain.Using the idea of dimensionality reduction,the maximum slope direction is converted into a two-dimensional profile,and the curve fitting method is used to obtain the characteristic change curve of the slope terrain,and the complete point cloud data of steep slope terrain is extracted.(5)Experiment with different terrain data and analyze the experimental results.The results show that the terrain classification can improve the applicability of complex point cloud data of different terrains.The algorithm uses adaptive threshold parameters for filtering and has certain practical value.