Blind identification of linear degradation operators in the Fourier domain

Abstract

Captured real images are often affected by environmental and technical interferences such as linear homogeneous distortion, nonuniform illumination, sensors' noise, geometrical scene distortion, etc. Among these issues, the former is particularly interesting because various physical problems can be modeled by such degradations. In this work we propose a blind algorithm for identification of the linear distortion operator based on the analysis of zero-crossings and phase distribution of the distorted image spectrum. The algorithm is tested with common real linear distortion operators and its identification rate is discussed.