The Management And Visual Application Of Massive Airborne LiDAR Point Clouds Data

LiDAR is a new technique for rapid access to high precision3D information of ground and terrain, the system acquires a series of three-dimensional point cloud data which is discrete spatial distribution of irregular. There are some urgent solution problems:how to manage massive point cloud data; how to display by using the method of dynamic scheduling; how to extract the information of terrain and surface features; how to apply the point cloud data in the field of topographic mapping, city construction and forestry planning. This paper focuses on:how to manage the massive laser point cloud data and display with dynamic scheduling; how to extract the ground points of interest target area and to extract the points of interest buildings in model of dynamic scheduling and display. This paper do the following research:First:This paper proposes the method of LOD based on virtue grid to achieve the management and dynamic scheduling display of massive airborne laser point cloud data. Firstly, using the method of virtual grid to block the surveyed area data, it can preserve the original data and also will not affect the data accuracy; Secondly, In order to facilitate subsequent dynamic scheduling, this paper use some methods to layering each block data, writing the layers data into files and recording the starting position of each layer to forming index files; Lastly, This paper completes the dynamic scheduling using the index file and data file and adopts OpenGL technology to show the point cloud data.Second:Based on the dynamic display, the users can observe the terrain of the interesting block, so they can set appropriate parameters to extract the ground points, which can enhance the adaptability of the algorithm, when executing the ground points, the program dynamically constructs the K-D tree of the interesting block to accelerate the speed of querying the field of the point.Third:Extracting building points interactively. Because of this paper research on extracting building points based on dynamic scheduling display, therefore, operators need to select a seed point from each target building, then according to certain rules to automatically extract the points of the building, finally completing the building contour constructed and subsequent modeling of three-dimensional buildings.