Lidar-based Coastline Extraction And Nature Recognition

The coastline,as the dividing line between land and sea,is one of the most important topographical basic elements on topographic and sea charts.Therefore,a comprehensive,efficient and accurate determination of the location,nature and dynamic changes of the coastline is of great practical significance for economic and national defense construction.LiDAR(Light Detection and Ranging,LiDAR)technology is another technological revolution in the field of surveying and mapping since GPS technology.Its emergence has brought new research hotspots to coastline surveying.It is compatible with artificial field surveys and photogrammetry.Complementary technical advantages.Based on summarizing the various methods and technologies of current coastline extraction,this paper focuses on the research of coastline extraction and property recognition technology based on LiDAR point cloud data.The main work and innovation are as follows:1.A systematic comparative analysis of LiDAR technology and other remote sensing technologies for coastline surveying,and a comparative analysis of the advantages of LiDAR technology for coastline surveying in my country in terms of platform,detection mechanism,waveband,etc.,and discussing the development of LiDAR technology for coastline surveying in my country Feasibility and necessity.2.Aiming at the problems of large loss of accuracy and jitter in flat beaches in coastline extraction from point cloud data,a coastline extraction method with multiple constraints of coarse and fine grids is proposed.This method uses the Mean High Water Springs(MHWS)as the segmentation threshold,divides the point cloud data into sea and land,and divides the part of the land into a coarse and fine grid,and uses edge extraction,expansion and corrosion,and hole filling And connected area detection combined processing to achieve coastline extraction.Through three coastline extraction experiments in different regions,the reliability and effectiveness of the coastline extraction method under multiple constraints of coarse and fine grids are verified.3.In view of the problem that the current researches on coastline extraction at home and abroad are too purely focused on the extraction of location lines and have not confirmed their nature information,this paper proposes a method for automatically identifying the nature of coastlines,that is,the random forest algorithm is used to first determine the coastal zone.The topographic point cloud is qualitatively classified.On this basis,the coastline extraction method under multiple constraints of the coarse and fine grids proposed in the article is used to obtain the point data of the coastline fragments with attributes,and the attribute integration is carried out incombination with the "Chinese Navigational Chart Compilation Specification".Through comparative analysis with artificially marked coastlines,this method can accurately identify coastlines of different natures,verifying the feasibility of this idea.