Design Of System For Lidar Data Processing And Tree Parameters Extracting

Terrestrial 3D laser scanner has the characteristics of convenience and simplicity, high-precision and high-efficiency in the acquisition of geometrical parameters of tree point cloud, so the research and application of LiDAR point cloud of trees have attracted much attention. But currently, there are few mature software systems which can integrate visualization and point cloud processing algorithm well, and there are fewer systems which can deal with tree point cloud processing and analysis. In view of this situation, on the basis of intense research on.NET development platform and Direct X11.0 3D visualization engine, this paper focuses on LiDAR point cloud processing and tree parameters extracting problems, and discusses traditional and improved algorithms of point cloud registration and estimation algorithm of part geometrical parameters. At last, through integrating 3D visualization interactive operation, data processing, tree parameters extracting algorithm and basic functions, the system for processing tree point cloud and extracting parameters is designed and developed. Main results and conclusions are as follows:(1) By improving the ICP algorithm based on K D tree for registration, we have achieved a fast and accurate registration method of trees point cloud. Through discussion of coarse registration method based on singular value decomposition and Target Center pinpoint registration method, the experiment shows that there are certain deficiencies in the accuracy and the applaction range of the two methods. The article presents an improved ICP algorithm based on KD tree for fine registration of tree point cloud. The experiment shows that compared with traditional ICP algorithm, the improved algorithm has an advantage in terms of precision and time.(2) On the basis of analysis on point cloud 3D visualization and interactive operation theory, we chose.NET as development platform frame DirectX11.0 as the 3D visualization engine Visual Studio 2010 as the development environment and C# as the programming language to design and develop the systematic function modules. The primary functions of the system includes file option, camera transformation, view and projection transformation, point cloud coordinate pickup, point cloud selection, point cloud format transformation, center of rotation setting, point cloud coloring and corresponding bar graph and pie chart statistic analysis, bounding sphere displaying, point highlighting,data registration and parameters extracting(3) On the basis of interactive operation, this paper introduces the system step by step by using the experimental data. The results indicate that the system shows good performance of human-computer interactions and the basic 3D interactive features. At last, the system completes the fine registration of tree point data and accurate extraction of tree parameters on the basis of 3D interactive operation.