Research On High-resolution Satellite Images Of Road Extraction Method

Urban roads is the skeleton of the city, it is an important part of the city's basic geographic information, complete information on all aspects of urban roads are of great significance. Remote sensing of urban roads is also an important feature in recent years due to the continuous improvement of remote sensing image resolution, high-resolution remote sensing images for further study road extraction more practical significance. High-resolution satellite images for this article, based on the basic characteristics of the road in the image, the study of the relevant road extraction algorithm, has made several achievements following:(1) Algorithm for automatic extraction of mathematical morphology, initial contour extraction method for road studies:Firstly iteration of the road in general area OTSU segmentation algorithm directly from the image, and the use of mathematical morphology opening and closing operation of the segmentation result of small removal of the communication block, to fill large gap, and edges slightly smoothed, the outline road finally obtained, and then on the basis of the initial road boundary obtained by the method of mathematical morphology edge extraction.(2) Based on the feature points extracted polygonal approximation of thought, there is a very classic algorithms that Douglas-Peucker method, application of mathematical morphology method to extract DP curve fitting method on the basis of road boundary, a boundary line for each role to play smooth.(3) The mathematical morphology to extract the closed road boundary as the initial contour Snakes model, the boundary line of the road were closed area iterative optimization, non-enclosed area treated with a polygonal approximation method, finally get a smoother road boundary, thus further improving extraction effect.