Super-resolution image restoration by combining incremental wiener filter and space-adaptive regularization

Abstract

Space-adaptive regularization (SAR) is a method for image restoration and it can decrease ring artifacts arising around sharp intensity transition of the image. The approach proposed in this paper incorporates incremental wiener filter (IWF) with SAR by edge-picking method and then estimates a single high-resolution image from different low-resolution images. The characters of spatial piecewise smoothness in SAR and rapid convergence in incremental wiener filter are combined in the new approach. The computer simulations show that this method has good convergence performance.