Quality Evaluation Algorithm: New Structural Similarity by Using Distance Transform Approach and Gradient Similarity
Zianou Ahmed Seghir, Fella Hachouf, Mounir Hemam, Frédéric Morain-Nicolier, Ahmed Zeggari · 2018
The image quality assessment (IQA) problem is treated inside this work, specially full-reference measure. Its objective is to deal with structural similarity (SSIM) index to treat the color images by introducing color distortion concept and distance transform (DT). First, the test and reference images are converted to gradient images using edge detection measure. Then the classical SSIM is computed using the gradient images, except, the luminance comparison. After that, the distance transform images are calculated by well know algorithm. The color distortion is determined by transforming images from RGB color space to YIQ. Finally, all previous results are combined to construct the hole error. Comparative study is made with measures using different types of distortions, lead us to say that the proposed method gives good results.