No-Reference Image Sharpness Assessment Based on Perceptually-Weighted Image Gradients
Loading...
Date
2023
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
IEEE
Abstract
The fact that human visual perception is highly spe-cialized for extracting structural information has been efficiently exploited in full-reference image quality assessment algorithms; additionally, it has been shown that the human perception of sharpness depends on the local image contrast. Since the image gradients can measure effectively the image structure, we propose a training-free no-reference objective image sharpness assessment method based on the statistical analysis of perceptually-weighted normalized -gradients of relevant pixels in the input image. Results over six subject-rated publicly available databases show that the proposed no-reference sharpness assessment algorithm correlates well with perceived sharpness and provides competitive performance when compared with state-of-the-art algorithms despite its low computational burden.
Resumen
Keywords
Image blur, Image sharpness, No-reference image quality assessment, Objective blur assessment, Perceptual-based