Applied Watershed Segmentation Algorithm for Water Body Extraction in Airborne SAR Image

Abstract:
A challenging task in computer vision and pattern recognition is digital image segmentation. Synthetic aperture radar (SAR) provides a unique view by measuring the radar backscatter response of the Earth’s surface. Watershed segmentation algorithm is well designed for image segmentation. However, it is excessively sensitive to speckle noise in SAR image which leads to oversegmentation that reduces efficiency. This paper presents a novel approach for water body extraction in SAR image by applying watershed combined with top-hat transformation. Formerly, spatial resolution constraints have limited the usefulness of SAR imagery for monitoring water bodies and extracting their edges. The purpose of this study is to improve the efficiency of morphological watershed techniques compared with Canny edge detection results for fine resolution (63cm) airborne SAR images and to yield an intuitive and well segmented image for mapping of water body boundaries.