Research On Airborne LiDAR Data Parallel Processing Algorithm For Disaster Emergency Response

In the emergency relief work,surveying and mapping geographic information plays an important role in basic guarantee and technical support.For this reason,many chapters in the national emergency plan for natural disaster rescue published by the general office of the State Council,such as disaster early warning response,national emergency response,post disaster rescue and recovery and reconstruction,define the responsibilities of the mapping geographic information department.Emergency mapping has the characteristics of responding to acute,sudden,flexible,special task organization,diversity of results and so on.These characteristics require us to break the traditional way and provide real-time and rapid mapping guarantee.However,the traditional surveying and mapping geographic information technology has many problems,such as large amount of data,many processing procedures which results in low efficiency of data post-processing and can not meet the requirements of emergency surveying and mapping for timeliness.In view of the above problems,this paper proposes a fast parallel processing method for Airborne LIDAR point cloud data.During preprocessing,coordinate transformation and filtering processing are carried out for point cloud data in combination with parallel algorithm,and block processing is carried out according to tile method,In the stage of generating DSM by point cloud interpolation,the direct interpolation algorithm based on parallel is used for interpolation,and in the stage of generating tile by DSM,the parallel computing is used for geographic information service.This method optimizes the processing flow of point cloud data,shortens the processing time and improves the processing efficiency,so that airborne laser point cloud data can play a full role in emergency mapping.In this paper,the airborne lidar scanning data with poor visualization is quickly processed into DSM data with strong visualization and easy processing,and applied to geographic information services to meet the requirements of emergency mapping.Through the experiment of traditional point cloud data processing and this method on Airborne LIDAR point cloud data in Guangyuan,Sichuan Province,the accuracy and efficiency of the results obtained by various methods are evaluated and analyzed,and the following conclusions are drawn:1.In terms of accuracy,the radius filter can effectively filter out the noise points caused by reflection and scattering of point cloud data in a short period of time,and improve the accuracy of point cloud data.Compared with the traditional point by point interpolation algorithm and the direct interpolation algorithm in the point cloud interpolation part,it is determined that the direct interpolation algorithm combined with parallel processing can meet the relevant requirements of emergency mapping in terms of accuracy compared with other interpolation algorithms.2.In terms of efficiency,compared with the traditional airborne lidar data processing methods,this paper improves the processing efficiency of large amount of point cloud data by using parallel computing and tile partition methods,and uses the method of direct interpolation.To a certain extent,the massive point cloud data exceeding the memory can be interpolated,greatly reducing the block interpolation processing time of traditional point cloud data processing,and saving the time cost in post-processing.3.In the face of emergency mapping,this paper optimizes the processing flow of Airborne LIDAR point cloud data,such as block,filter,interpolation and tile,to a certain extent,to achieve the rapid processing of point cloud data.At the same time,it improves the efficiency of processing large amount of Airborne LIDAR point cloud data to meet the time requirements of emergency mapping,and improves the application environment of airborne lidar scanning technology in emergency mapping.