The SAR image segmentation superpixel-based with optimized spatial information

Xiaolin Tian, Licheng Jiao, Long Yi, Xiaohua Zhang

Abstract: In this paper, we propose a method of image segmentation, which is based on superpixel and optimized spatial feature. In this paper, the superpixels are taken into account, which can reduce computational burden, and the result of over-segmentation can also be benefit for segmentation results. The main idea of this paper is based on the conventional fuzzy c-means (FCM). The conventional FCM has a better performance. However, it is sensitive to noise. In order to overcome this shortage, we incorporate spatial information of superpixels into the conventional FCM. In order to obtain the better performance, influential degree of spatial information is applied to the conventional FCM to improve segmentation performance Experimental results show that the proposed method achieves excellent performance.