Perceptually optimized blind repair of natural images

作者：

Highlights：

• A method to use perceptual image quality assessment predictions to perform distortion-blind and distortion-severity-blind image repair.

• A method to perceptually optimize the parameters of distortion repair algorithms so that the repaired image attains maximum perceptual quality.

• A general purpose framework that allows for iterative distortion repair to maximize output perceptual quality.

摘要

Highlights•A method to use perceptual image quality assessment predictions to perform distortion-blind and distortion-severity-blind image repair.•A method to perceptually optimize the parameters of distortion repair algorithms so that the repaired image attains maximum perceptual quality.•A general purpose framework that allows for iterative distortion repair to maximize output perceptual quality.

论文关键词：Image quality,Image repair,Distortion identification,Perceptual image repair

论文评审过程：Received 10 September 2012, Revised 8 July 2013, Accepted 21 August 2013, Available online 6 September 2013.

论文官网地址：https://doi.org/10.1016/j.image.2013.08.007