Unsupervised Blind Quality Estimation Of NSS Images And Repair Of Low Quality Images

Abstract

Present world works with digital images and digital content in each and every day in life. But the consideration behind it in the area of quality of the content is currently less and is increasingly becoming popular. Our work is mainly in the area of tossing light towards the Blind or No reference quality analysis of Natural Scene Statistics(NSS) images and to the possible extent repairing it after judging its quality. The study, experiment and analysis of the result is carried out with the help of distortion aware and opinion unaware NSS images collected from LIVE dataset and the results have shown achievement in the direction of proper quality estimation and correction of noisy images.