Two Coastline Detection Methods Based on Improved Level Set Algorithm in Synthetic Aperture Radar Images

Level Set Algorithm (LSA) is one of the efficient tools of edge detection and contour extraction, which has been applied to radar, medical, and other images processing successfully. However, the procedure using LSA will cost a lot of time when applied to high-resolution images. This paper provides two novel approaches that can enhance the detection speed. In the first method, the moving step enlarges when the contour is far from the real boundary of specific targets, while reduces when the contour is near the real boundary in the images. In the second method, reducing resolution make the coarse boundary be obtained by LSA in low resolution image. Then, the contour is processed finely by LSA in high resolution image based of the coarse boundary. Experimental results using Radarsat ScanSAR images show the effectiveness of these new methods and verify the improved velocity in contour detection compared with the standard LSA.