Research Of Road Extraction Based On Airborne LiDAR And High-Resolution Remote Sensing Imagery

Based on the single data source, the road center lines are extracted, while theextracted results are mainly the urban main roads, some narrower roads are ignored. Inthis thesis, the LiDAR (Light Detection and Ranging) point clouds data and high-resolution remote sensing data are associated to extract road information to improve theaccuracy. Firstly, the LiDAR point clouds data and high-resolution remote sensing dataare registrated by using ENVI, Matlab, ArcGIS and other softwares, combined withLiDAR data in the study area. And then road information is extracted by the fusionmethod and the Object Shape Index feature on the basis of image segmentation andimage object-based hierarchical extraction. At last, by superposition visual analysis andquantitative analysis, giving the accuracy evaluations and analysis for the LiDAR dataand high-resolution remote sensing combinative based extraction method.The main research results obtained in this thesis are as follows:(1) The basic theory of remote sensing image segmentation and feature extractionwas used to complete the work of the high resolution remote sensing imagesegmentation and object oriented road feature extraction;(2) A kind of object description of road shape: the Object Shape Index which wasused for the road surface extraction was put forward based on the segmentedimage;(3) Complete the software prototype development of the object-oriented featurecomputing and road extraction based on ArcEngine10.1.Studies have shown that the road information extraction method which integratesutilizing LiDAR point clouds data and high-resolution images has better results thansingle using remote sensing images. The proposed method in this thesis is feasible andeffective.