Ocean-land segmentation based on active contour model for SAR imagery

Ocean-land segmentation is a basis for oceanic targets detection for SAR imagery. The traditional manual segmentation method can acquire a fairly good result, while it is unable to achieve a satisfying efficiency. Thus the manual method is unsuitable for batch processing. Ocean map based land masking method requires a priori information, which restricts its application. This paper presents an automatic active contour model based ocean-land segmentation method for SAR imagery. This method forces the active contour to the real coastline by variation level set. Under the restriction of minimizing a given energy function, the method obtains the final segmentation result when the active contour is overlapped on the coastline. In addition, an optimizing method is proposed to improve its computation efficiency, which makes our segmentation method more applicable.