A full-reference stereoscopic image quality metric based on binocular energy and regression analysis

The recent developments of 3D media technology have brought to life numerous applications of interactive entertainment such as 3D cinema, 3DTV and gaming. However, due to the data intensive nature of 3D visual content, a number of research challenges have emerged. In order to optimise the end-to-end content life-cycle, from capture to processing and delivery, Quality of Experience (QoE) has become a major driving factor. This paper presents a human-centric approach to quality estimation of 3D visual content. A fullreference quality assessment method for stereoscopic images is proposed. It is based on a Human Visual System (HVS) model to estimate subjective scores of registered stereoscopic images subjected to compression losses. The model has been trained with four publicly available registered stereoscopic image databases and a fixed relationship between subjective scores and the model has been determined. The high correlation of the relationship over a large number of stimuli has proven its consistency over the state-of-the-art.