Foveation-based content Adaptive Structural Similarity index
Snježana Rimac-Drlje, Goran Martinović, Branka Zovko-Cihlar · International Conference on Systems, Signals and Image Processing · 2011
This paper proposes a novel video quality metric — Foveation-based content Adaptive Structural SIMilarity index — FA-SSIM, which combines the Structural SIMilarity index (SSIM) and the foveation-based sensitivity function S f . Function S f is based on the Human Visual System (HVS) behavior that spatial resolution is the highest around the point of fixation and decreases rapidly with increasing eccentricity. Furthermore, S f captures effect of further spatial acuity reduction due to motion in a video sequence. Accuracy of proposed metric is tested on the LIVE video quality database. The results of quality evaluation show that FA-SSIM outperforms SSIM as well as VQM and MS-SSIM metric and that is even competitive with state of the art MOVIE metric.