PROJECT TITLE :

Corrupted Reference Image Quality Assessment of Denoised Images

ABSTRACT:

As a new foundation for understanding picture quality and denoising, we present corrupted reference image quality assessment (CRIQA). Denoised images are compared to ideal reference images (not provided) by predicting the full-reference image quality assessment (FRIQA) scores of denoised images. Simulated and human subject tests show that the CRIQA scores of denoised images coincide with the corresponding FRIQA scores, and these scores are more consistent with the perceived denoising quality than NRIQA scores. CRIQA's value was proved in a denoising parameter adjustment application.