A Method of Road Extraction from High Resolution Remote Image Based on Delaunay Algorithms

摘要

Aimed at the problem of the insufficient use of road features in traditional road extraction methods. To overcome the limitations, a new road extraction method from high resolution satellite image based on Delaunay algorithms was proposed in this paper. Firstly, the images should be pretreated by edge detection and binarization, etc. Secondly, the Delaunay triangulations were constructed according to the Delaunay triangulation algorithm, which could represent the image information. Each triangle in Delaunay triangulation was defined as the basic processing unit, namely triangle-unit. Thirdly, combined with the road features, the four feature parameters were put forward after analyzing and summarizing the characters of triangle-units. Then, the triangle-units belonging to road section were extracted depending on the feature parameters. Finally, road edges and road centerlines were automatically extracted by road triangle-units. In this paper, the feasibility of the method was verified by the road extraction experiment of UAV images.