Road Extraction from Multi-source Remote Sensing Images Based on 2nd Generation Curvelet Fusion

In view of the problem of non-target interference and misjudgment in road extraction from remote sensing images, a new method of road extraction based on curvelet fusion is presented. Firstly spectral similarity matrix is calculated in multispectral images. Then the panchromatic image and the spectral similarity matrix are fused by using curvelet transform. The result is processed with cross-correlation line detection and isolated pixels suppression. Finally the road features are extracted by using Hough transform. In order to verify the effectiveness of the algorithm, experiments are performed in many different scenes. The results show that the method has good robustness and accuracy. Roads can be extracted accurately in both simple and complex scenes.