Speckle suppression in synthetic aperture radar ocean internal solitary wave images with curvelet transform

Abstract

This paper proposes a speckle-suppression method for ocean internal solitary wave (ISW) synthetic aperture radar (SAR) images by using the curvelet transform. The band-shaped signatures of ocean ISWs in SAR images show obvious scale and directional characteristics. The curvelet transform possesses a very high scale and directional sensitivity. Therefore, the curvelet transform is very efficient in analyzing wave signals in SAR images. A noisy ocean ISW SAR image can be decomposed at different scales, directions, and positions using the curvelet transform. The information of the ISWs is centralized in the curvelet coefficients of certain directions under certain scales, whereas the speckle noise is distributed in every scale and direction. By manipulating the curvelet coefficients, the signals of the ISWs can be extracted from the noisy SAR image. Finally, the speckle noise is suppressed and the ISW feature is enhanced by adding the signals of the ISWs back to the original SAR image. Experiments demonstrate the effectiveness of this method.