A reduced reference distortion estimation measure for color images

Estimation of distortion in images using reduced reference mode of quality assessment is a reliable approach to develop quality metrics which are in due coherence with human visual system (HVS). This paper proposes a novel reduced reference image quality assessment methodology employing multi-resolution approaches for features extraction. The extracted features are then used to formulate a distortion measure as a function of contrast, luminance and divergence factors. The derived distortion measure serves as a versatile tool to predict the image quality across several distortion types. Simulation results depict larger coherence between the evaluations performed by proposed distortion measure on images from CSIQ database.