Airport runway detection based on improved fuzzy clustering for PolSAR images

This paper proposes a fast airport runway detection method for PolSAR image based on improved fuzzy C-mean method (IFCM). To avoid the issue that FCM is sensitive to random selected class centers, polarimetric scattering characteristics of the targets in PolSAR image are used to generate appropriate initial centers. And a simplified Wishart distance is adopted instead of Euclidean distance in order to obtain a fine classification result. Then, regions of interest which contain runway are automatically extracted from the classification result with the weak scattering characteristic of airport runway. Finally, structural features are used to identify real runway area. Experimental results show that compared with the other two algorithms, the proposed method can detect the airport runway area with more completed structure and less computation time.