A combined first and fractional order regularization method for mixed Poisson-White spike noisy image restoration (Q6066509)

The problem of image restoration or image denoising is a fundamental area of research in computer vision. The problem of noise contamination occurs in many applications for example in neutron radiography in medical imaging. Given the importance of the areas in which it occurs, many methods have been developed over many years to deal with this contamination and ways to recreate the image without it. The paper under review deals with this interesting subject and many references are given for this subject. More specifically, the authors study the problem of restoring images effected by mixed Poisson-Square Cauchy noise. In order to do this, the authors derive a multi-convex variational model where a modified four-directional fractional order total variation is combined with a first order total variation to characterize the image prior. The authors derive a block coefficient descent based algorithm in which variable splitting and alternating direction minimization of multipliers are utilized along with anisotropic diffusion and additive operator splitting. Numerical experiments are given comparing their method with two other total variation regularized models with different fidelities which confirm the validity and flexibility of their method. The paper is well written with a set of good references.