Research On The Construction Of Complex Terrain DEM Based On LiDAR Point Cloud

Airborne LiDAR(Light Detection and Ranging,LiDAR)integrates laser ranging system,inertial navigation system,global positioning system and other technologies.It is a new type of active earth observation technology,and it is also a new technology in the measurement field after GPS.The revolution in scientific and technological development.It breaks through the traditional point-by-point measurement method,makes the measurement work present the characteristics of fast,efficient,and large-scale,so that the obtained LiDAR data has spatial three-dimensional information,and to a certain extent solves the inconvenience of traditional photogrammetry to obtain three-dimensional information.Airborne LiDAR measurement technology is widely used to realize the rapid collection of various large-area three-dimensional terrain data and the rapid generation of various digital terrain products.Digital elevation models are widely used in fields such as hydrology,meteorology,geology,military,and national defense construction.In particular,they can analyze landforms by factors such as slope,aspect,and slope conversion rate.This article relies on a Guangxi innovation-driven development project(major science and technology project)(subtopic five: integrated data processing and demonstration applications).According to the actual needs of the subject this paper studies the key technology of Lidar point cloud data generation digital elevation model,and introduces it from the following three aspects,point cloud data preprocessing,point cloud data filtering processing and point cloud data interpolation processing.The main research contents are as follows:1.Introduced the airborne lidar technology and point cloud data management methods.Starting from the composition of airborne lidar technology,the working principle of the airborne lidar measurement system is studied.The format characteristics of airborne lidar point cloud data are studied,and then the technical process of point cloud data preprocessing is introduced,and an improved K-nearest neighbor distance algorithm based on the K-D tree structure is proposed to remove the noise in the point cloud data.2.Research the algorithm of point cloud data filtering.Starting from the principle of lidar point cloud data filtering,the point cloud data distribution characteristics are studied.Point cloud data filtering The traditional algorithm for filtering a single terrain feature is better,but the filtering effect is not very good for complex terrain.Therefore,this paper proposes a cloth simulation filtering algorithm that takes into account the terrain,so that the filtered ground point cloud data has data The scene information and the structural characteristics of the terrain.3.Research the point cloud data interpolation algorithm.The data structure format of the digital elevation model is introduced.Starting from the principle of the interpolation algorithm for lidar point cloud data,the characteristics of the interpolation algorithm for point cloud data are introduced.Recalling the classic research on traditional point cloud data interpolation algorithms,a cutting-edge advancement algorithm based on TIN was proposed to build a highprecision digital elevation model.