A study of human perception similarity for image quality assessment

Subjective image quality evaluation, though time consuming, is presently the most reliable evaluation method. The aim of this research was to undertake an independent study to investigate the similarity between a psychophysical experiment and an existing image quality database, where the experiments were undertaken in geographically distant locations with racially dissimilar population and varied number of subjects. The image quality database obtained from Laboratory for Image and Video Engineering (LIVE) was used to carry out subjective evaluations. The Difference Mean Opinion Scores (DMOS) were calculated from the raw scores and were analyzed against those from LIVE. The results indicate a high correlation between the two evaluations, thus confirming a globally consistent human perception.