Research On Extraction Technology Of Land Cover Information Based On Stereo Image And Airborne LiDAR Data

With the institutional reform and business readjustment of the ministry of natural resources,the monitoring of geographical conditions has become an important part of the survey and monitoring of natural resources.However,all kinds of natural resource management services are micro-scale and specific,rather than the macro and comprehensive information needs of the previous geographical situation monitoring.With the change of application demand,new requirements have been put forward for the content index and technical route of the geographical condition monitoring,especially for the spatial positioning accuracy and content refinement accuracy of the surface cover map.The traditional information extraction of surface coverage spot monitoring is based on the single-field orthography.Although the single-field orthography has correct plane information and rich image information,it lacks the elevation information of surface coverage and cannot improve the collection accuracy and interpretation accuracy of surface coverage.Both stereo image and airborne LiDAR data have rich elevation information.This paper mainly studies the use of stereo image and airborne LiDAR as data sources to extract the classification information of geographical conditions.This experiment combined with the distribution of data sources,and xunyi county,fu county of shaanxi province by the area as the research object,make full use of the land types elevation information in the data source,focused on cultivated land,grassland,shrubbery,multilayer and more independent buildings the features of easy to mix in the surface cover type,such as,type of information extraction experiments.For stereo images,the experiments of satellite stereo images preprocessing steps for comb,for different image source in aerial triangulation link was studied,respectively under the stereo images of different resolution of surface coverage,this paper compares and analyzes the effect of information extraction in the process of experiment as the earth class boundary acquisition precision and class factors had significant effect on the interpretation accuracy.For airborne LiDAR,experiment exploring point cloud data processing methods,steps and summarizes the using point cloud data in remote sensing image and the combination of joint interpretation key establishing typical classification,based on information extraction,and analyzed the result of the extraction,and explores the differences between cloud density on the result of extraction,trial production for the future demand for point cloud data quality are proposed.Study by experiment,the different data sources is adopted to improve the surface coverage of information extraction experiment technical route,for the future in geographic conditions monitoring project using stereo images and airborne LiDAR data source to carry out the trial production accumulated technology,put forward feasible technical support,and for the technology to lay a good foundation for engineering application in the country.