作者: Vittoria Bruni , Domenico Vitulano
DOI: 10.1109/ISPA.2013.6703799
关键词:
摘要: This paper investigates the properties of Jensen-Shannon divergence for visual quality assessment. It will be shown that is a potential adaptive tool thanks to good correlation with human perception, reduced computational effort and limited amount information on original image. In addition, it gives score in terms bits has theoretical relationships some well known highly performing image assessment measures. Preliminary experimental results show proposed measure predicting degradation analysed valid feature successfully use preservation restoration archived material.