Indices for image quality degradation evaluation based on wavelet transforms

Teruya Minamoto, Hajime Ohmura · 2014

We present new indices for measuring the degradation of a digital image. Objective measures for assessing image quality, such as PSNR and SSIM, have been developed to quantify the visibility of errors or structural information between the original image and its modified version. However, the results obtained by PSNR and SSIM do not always match a subjective evaluation. We developed new indices based on the continuous wavelet transform (CWT) for measuring degradation so as to match a subjective evaluation. We also developed some indices based on the discrete wavelet transform (DWT) and investigated the usefulness of the proposed indices by conducting experiments.

Read the paper · More papers on PaperTik