Article ID Journal Published Year Pages File Type
4970437 Signal Processing: Image Communication 2017 14 Pages PDF
Abstract
Quality of Experience (QoE) measurement in stereoscopic 3D is a challenging topic and existing standards such as ITU-R BT.500 may no longer be adequate for reliability and efficiency. Yet, objective measurement of QoE is not well researched for 3D images and videos. We present a new subjective assessment methodology providing better time-efficiency and more accurate scores at the same or even less number of assessors, using new objective metrics for QoE measurement of compressed multi-view video. Our research focus is twofold. First, the novel subjective testing methodology is developed, and QoE of different multi-view video compression technologies is evaluated. A time-efficient solution is proposed as a multiple stimuli plus simultaneous presentation extension of the conventional Double Stimulus Continuous Quality Scale (DSCQS) method. Second, objective quality metrics are proposed based on structural similarity and depth maps. Experiments are conducted on an auto-stereoscopic display while H.264/MPEG-4 Advanced Video Coding (AVC)- and High Efficiency Video Coding (HEVC)-based encoders are used for compression. The results demonstrate that the new methodology is superior to the DSCQS method in terms of both evaluation time and accuracy. Furthermore, high correlation is achieved with the new objective quality metrics.
Related Topics
Physical Sciences and Engineering Computer Science Computer Vision and Pattern Recognition
Authors
, ,