Objective Reduced reference Stereoscopic image quality assessment

Abstract

In the last years, 3D imaging is becoming a vast area of research focused on several disciplines such as industry and scientific applications. Nevertheless, the evaluation of 3D quality does not follow the technical evolution. This paper investigates the evaluation of the quality of stereoscopic 3D images. Although several methods dedicated to 2D quality assessment have been presented in the literature, there was not as much effort for evaluating the quality of 3D images. It’s in this perspective and in the aim to introduce an objective method for evaluating the quality of 3D images, we suggest to combine the estimation of each 2D quality assessment to extend it to 3D case. Thus different metric have been tested on LIVE and IRCCyNIVC databases, the obtained results show that reduced reference methods are the most appropriate.