On Solving SAR Imaging Inverse Problems Using Nonconvex Regularization With a Cauchy-Based Penalty

Exploiting the Dual-Tree Complex Wavelet Transform for Ship Wake Detection in SAR Imagery

“…In addition to its role in modelling the heavy-tailed noise in various signal and image processing applications, Cauchy based penalty function has rarely been used as the signal prior in the literature. Based on the derivation of its proximal operator in [10], the work in [11] proposed the Cauchy penalty function for addressing four SAR imaging inverse problems: super-resolution, image reconstruction, despeckling and ship wake detection. The Cauchy based penalty function has shown improved performance over various important penalty functions in [11] such as L1, TV, and GMC.…”

“…Although some efforts on ship wake detection based on inverse problem solving have been done by employing penalty functions such as GMC [8], and Cauchy [11], both penalty functions are in need of further improvements mainly for better tackling noise artefacts, which sometimes induce false wake detections. DT-CWT appears a promising tool for efficiently enhancing the corresponding SAR image for the purpose of wake detection, thanks to its properties of reducing the translation sensitivity and improving the directional selectivity [18].…”

“…the ensuing algorithm guaranteed to convergence [10]. This led to higher ship wake detection accuracy in [11,12] compared to the GMC, L1 and total variation (TV) penalties.…”

“…In this paper, we propose a novel algorithm for detecting ship wakes, which is based on imaging inverse problems to enhance linear features in SAR imagery. We achieve this by combining the non-convex Cauchy based penalty function [11] with the DT-CWT. The proposed method, which benefits from the near shift-invariance property of the DT-CWT, along with the heavy-tailed nature of the Cauchy penalty function, is then compared to methods based on several state-of-the-art penalty functions, such as GMC [7], T V , Cauchy [11], and the method of Graziano et.…”

SAR image denoising based on multifractal feature analysis and TV regularisation

“…Multiple speckle denoising techniques have been proposed in the literature [1][2][3][4][5][6][7][8][9][10][11][12][13][14][15][16][17][18][19] which aims to restore speckle corrupted images. To begin with, a multiplicative noise filtering model was proposed in [20].…”