Blind restoration and resolution enhancement of images based on complex filtering

Abstract

Recently, a blind resolution enhancement method that uses a two-dimensional and single-input multiple-output extension of the constant modulus algorithm has been developed for pure translational motion. The method works well in case of low bit depth unobserved true images, but its performance decreases for high bit depth true images. In this work, we propose a refined scheme in which complex representation of images and a set of complex deconvolution FIR filters are used. Simulations show that the refined method succeeds in reconstructing the low and high bit depth true images without the knowledge of blur parameters. Visual results for the restoration case (single image, no subsampling) are also given. No assumption is made about the blurs except that they have low-pass characteristics. Also, they do not have to be the same for the observed low-resolution images and they do not need to be shift invariant.