Multiplicative denoising based on linearized alternating direction method using discrepancy function constraint (Q474952)

This paper considers the problem of recovering the original image from the observed image contaminated by a multiplicative noise. The objective function is a weighted sum of two terms: the 1-divergence as the data-fidelity term and the total variation regularization term. The authors develop a linearized alternating direction minimization algorithm with auto-adjusting of the regularization parameter for the gamma noise removal. In each iteration, one quadratic term of the augmented Lagrangian function is linearized and the regularization parameter is updated to satisfy the convex discrepancy function constraint. The global convergence of the algorithm is established. The efficiency of the proposed algorithm is illustrated by some numerical results.