No-reference image quality assessment in curvelet domain

作者：

Highlights：

• CurveletQA exploits a model of the log-pdf of curvelet coefficients to find the statistical correlations between curvelet scalar and orientation energy distributions and image distortions.

• CurveletQA correlates highly with the human subjective impressions and has a relatively low time complexity.

摘要

•CurveletQA exploits a model of the log-pdf of curvelet coefficients to find the statistical correlations between curvelet scalar and orientation energy distributions and image distortions.•CurveletQA correlates highly with the human subjective impressions and has a relatively low time complexity.

论文关键词：Image quality assessment (IQA),No reference (NR),Curvelet,Natural scene statistics (NSS),Support Vector Machine (SVM)

论文评审过程：Received 1 August 2013, Revised 24 February 2014, Accepted 24 February 2014, Available online 6 March 2014.

论文官网地址：https://doi.org/10.1016/j.image.2014.02.004