For Full-Text PDF, please login, if you are a member of IEICE,|
or go to Pay Per View on menu list, if you are a nonmember of IEICE.
Objective No-Reference Video Quality Assessment Method Based on Spatio-Temporal Pixel Analysis
Wyllian B. da SILVA Keiko V. O. FONSECA Alexandre de A. P. POHL
IEICE TRANSACTIONS on Information and Systems
Publication Date: 2015/07/01
Online ISSN: 1745-1361
Type of Manuscript: PAPER
Category: Image Processing and Video Processing
H.264, MPEG-2, no-reference metric, objective video quality assessment,
Full Text: PDF>>
Digital video signals are subject to several distortions due to compression processes, transmission over noisy channels or video processing. Therefore, the video quality evaluation has become a necessity for broadcasters and content providers interested in offering a high video quality to the customers. Thus, an objective no-reference video quality assessment metric is proposed based on the sigmoid model using spatial-temporal features weighted by parameters obtained through the solution of a nonlinear least squares problem using the Levenberg-Marquardt algorithm. Experimental results show that when it is applied to MPEG-2 streams our method presents better linearity than full-reference metrics, and its performance is close to that achieved with full-reference metrics for H.264 streams.