Multiscale contrast similarity deviation: An effective and efficient index for perceptual image quality assessment
作者：
Highlights：
• Generally, the effectiveness, namely, high correlation with the human subjective score, is the prerequisite of a good IQA model, it is of first importance to a IQA model.
• The efficiency (the least computation cost), however, is the second requirement of a good IQA model which become important under the premise that an IQA model meet the condition of "effectiveness".
• In this paper, we developed an effective and efficient IQA model called multiscale contrast similarity deviation (MCSD) which explores the contrast features by resorting to multiscale representation.
• Performances on six benchmark databases demonstrate its effectiveness and efficiency.
摘要
•Generally, the effectiveness, namely, high correlation with the human subjective score, is the prerequisite of a good IQA model, it is of first importance to a IQA model.•The efficiency (the least computation cost), however, is the second requirement of a good IQA model which become important under the premise that an IQA model meet the condition of "effectiveness".•In this paper, we developed an effective and efficient IQA model called multiscale contrast similarity deviation (MCSD) which explores the contrast features by resorting to multiscale representation.•Performances on six benchmark databases demonstrate its effectiveness and efficiency.
论文关键词：Contrast similarity,Image quality assessment,Multiscale,Standard deviation pooling,Full reference
论文评审过程：Received 24 November 2015, Revised 20 April 2016, Accepted 20 April 2016, Available online 21 April 2016, Version of Record 27 April 2016.
论文官网地址：https://doi.org/10.1016/j.image.2016.04.005
