Blind image quality assessment using a reciprocal singular value curve

作者:

Highlights:

• We found out relationship between image distortion and reciprocal singular value curve.

• We constructed two new general blind image quality assessment (IQA) indices that respectively use the area and curvature of image reciprocal singular value curves.

• The proposed indices have the following advantages. (1) Simple mathematical expression leads to low computational complexity; (2) they can be applied to more distorted categories, such as “High frequency noise,” and “WN-color.”

摘要

•We found out relationship between image distortion and reciprocal singular value curve.•We constructed two new general blind image quality assessment (IQA) indices that respectively use the area and curvature of image reciprocal singular value curves.•The proposed indices have the following advantages. (1) Simple mathematical expression leads to low computational complexity; (2) they can be applied to more distorted categories, such as “High frequency noise,” and “WN-color.”

论文关键词:Image quality assessment,Singular value decomposition (SVD),Reciprocal singular value curve,No-Reference

论文评审过程:Received 1 November 2013, Revised 12 September 2014, Accepted 17 September 2014, Available online 28 September 2014.

论文官网地址:https://doi.org/10.1016/j.image.2014.09.005