An efficient human visual system based quality metric for 3D video

Abstract

Stereoscopic video technologies have been introduced to the consumer market in the past few years. A key factor in designing a 3D system is to understand how different visual cues and distortions affect the perceptual quality of stereoscopic video. The ultimate way to assess 3D video quality is through subjective tests. However, subjective evaluation is time consuming, expensive, and in some cases not possible. The other solution is developing objective quality metrics, which attempt to model the Human Visual System (HVS) in order to assess perceptual quality. Although several 2D quality metrics have been proposed for still images and videos, in the case of 3D efforts are only at the initial stages. In this paper, we propose a new full-reference quality metric for 3D content. Our method mimics HVS by fusing information of both the left and right views to construct the cyclopean view, as well as taking to account the sensitivity of HVS to contrast and the disparity of the views. In addition, a temporal pooling strategy is utilized to address the effect of temporal variations of the quality in the video. Performance evaluations showed that our 3D quality metric quantifies quality degradation caused by several representative types of distortions very accurately, with Pearson correlation coefficient of 90.8 %, a competitive performance compared to the state-of-the-art 3D quality metrics.Keywords 3D TV; stereoscopic video; depth map; cyclopean view; quality of experience; 3D video quality metric.

Saliency inspired quality assessment of stereoscopic 3D video

“…After collecting the subjective quality scores, an overall Mean Opinion Score (MOS) is calculated by averaging the individual subjective scores and removing the outliers [49]. An objective metric value is also computed using each of the FR and NR metrics to compare with the subjective results.…”

“…Human Visual system based 3D video quality metric (HV3D) evaluates the perceived 3D quality of stereoscopic videos as a combination of depth map quality and quality of the views. HV3D is formulated as a combination of three terms [49]:…”

“…For the third component (variances of the blocks), a single average saliency value (average of saliency map values) for each block is used as a weight for the variance term. The same constant parameter values are used as the ones reported in [49].…”

Optimized Adaptive Encoding Based on Visual Attention

“…These existing methods have been successfully applied to video signal redundancy processing, whereas few studies have taken note of visual redundancy. The color, brightness, contrast, and texture in 3D animation are robust; thus, visual redundancy research is particularly important [17]. This study proposes a novel adaptive encoding algorithm that is based on visual attention.…”

Full-Reference Quality Assessment of Stereoscopic Images by Learning Binocular Receptive Field Properties

“…Lee and Lee [28] proposed a 3D Perception-based Stereo image quality pooling (3DPS) model, which divides a stereo image into binocular and monocular vision segments, and evaluated them respectively. Banitalebi-Dehkordi et al [29] fused both left and right views to construct a "Cyclopean" view, and considered the sensitivity of the human visual system (HVS) to contrast and the disparity of the views. Jin et al [30] quantified 3D image quality into three components: cyclopean view, binocular rivalry and the scene geometry components.…”