A level set based method for coastline detection of SAR images

Abstract

Coastline detection is important for surveying and mapping reasons. This paper presents an efficient approach to detect coastlines from synthetic aperture radar (SAR) images. The proposed approach is based on fuzzy c-means (FCM) clustering and level set segmentation. It consists of a sequence of image processing algorithms. First, the FCM clustering is applied to the input SAR image. Second, the level set method has been used the result of FCM clustering as initial contours to extract the coastline. The method proposed in this paper, does not require determining the initial shape for active contour. Also it is robust to speckle noise. Experimental results on high and low resolution SAR images show the good performance of this method for coastline detection.