Prediction of visual fatigue from spatiotemporal characteristics in stereoscopic video

Abstract

Along with the increasing demand of 3D technology, the visual fatigue issue from stereoscopic video comes to the fore, which has been actively discussed in literature. Nevertheless, it is difficult to find a definite prediction model of visual fatigue in stereoscopic video due to the complexity of the human visual system (HVS). In this paper, we analyze the spatiotemporal characteristics such as depth, spatial frequency and motion. In addition, the human factor like the zone of comfort is also included to predict the visual fatigue. Consequently, we propose a novel prediction model by integrating the weights of several characteristics in stereoscopic video and verify its reliability through a subjective assessment.