Road Extraction by Point-wise Gaussian Models

We present a Gaussain model based approach for robust and automatic extraction of roads from very lowresolution satellite imagery. First, the input image is ltered to suppress the regions that the likelihood of existing a road pixel is low. Then, the road magnitude and orientation are computed by evaluating the responses from a quadruple orthogonal line lter set. A mapping from the line domain to the vector domain is used to determine the line strength and orientation for each point. A Gaussian model is tted to point and matching models are updated recursively. The iterative process consists of nding the connected road points, fusing them with the previous image, passing them through the directional line lter set and computing new magnitudes and orientations. The road segments are updated at each iteration, and the process continues until there are no further changes in the roads extracted. Experimental results demonstrate the success of the proposed algorithm.