Road network detection from SPOT satellite image using Hough transform and optimal search

Road networks are important features of satellite imagery. In this paper, we propose an algorithm for detecting a major road network from a satellite image. Firstly, a binary image representing edges of a given image is used to evaluate the possibility for a road to pass on each edge pixel. We propose an 8-direction filter, a clearing filter, and a parallel-edge-detection filter, which can compile insufficient local information to obtain sufficient global information for detecting a road. After employing these filters, we divide the image obtained into small images and detect road segments in small sub-images by the Hough transform. Then an optimal search method is employed to detect a global road and in turn, road pixels. This detection theory does not depend on the size of image and can detect almost all the main roads in an image including intersecting and T-type roads.