11-04-2017, 12:39 PM
A family of 3-D block adaptation algorithms (BM3D) has recently been proposed for various image problems, in the framework of non-local image modeling. In this work, we construct frames of analysis and synthesis, formalizing BM3D image modeling, and use these frameworks to develop new iterative deblurring algorithms. We consider two different formulations of the deblurring problem, ie one given by the minimization of the single objective function and another based on the generalized Nash equilibrium (GNE) of two objective functions. The latter results in the algorithm in which the deflating and dismantling operations are decoupled. The convergence of developed algorithms is demonstrated. The simulation experiments show that the decoupled algorithm derived from the GNE formulation demonstrates the best numerical and visual results and shows superiority with respect to the state of the art in the field, confirming a valuable potential of BM3D frames as an advanced image modeling tool .
It can be understood in the following video: