A-priori Information Driven Model for Road Segmentation in High Resolution Images

The problem of road segmentation in high resolution images is addressed in this paper. We present an extension of the active contour model that includes a-priori knowledge about the width of the roads being extracted. In order to improve the segmentation performance of the algorithm, the proposed model also contains a modified external energy term. The problem of the contour energy minimization is solved by using genetic algorithms, which makes it possible to use random initialization of the contours. The proposed segmentation method is validated on a series of high resolution images. The results show that by using the modified external energy term the new model performs better than the typically used gradient-based version.