Perceptual quality prediction on stereoscopic three-dimensional images using efficient local quality algorithms

Abstract

Stereoscopic image quality assessment (SIQA) is an important task in many applications. An efficient SIQA model should not only deliver high-quality prediction accuracy but also be computationally efficient. We propose a simple but very efficient quality assessment index to evaluate image quality using the global variation of image gradient magnitude. The quality assessment index is computed by modeling the perceptual gradient attributes of the reference stereoscopic images and the distorted stereoscopic images based on the deviation pooling strategy. Experimental results demonstrate that the proposed algorithm can serve as an efficient predictive image quality feature, which not only delivers highly competitive prediction accuracy, but also low computational complexity.