Efficient Models for Objective Video Quality Assessment

Abstract

Two models for objective assessment of compressed video quality and results from subjective and objective tests of several low bit-rate video coders are introduced in this paper. First model is based on mathematical measures that describe perception of video distortion by human eye. Second model for quality evaluation is based on human visual system (HVS) characteristics. Obtained quality values from both models have been compared with subjective results. Both models are computationally efficient and produce results that are correlated with subjective results.