Pixel Club Seminar: Quaternion Structural Similarity - a new Quality Index for Color Images

אמיר קולמן (הנ' חשמל והנ' מחשבים, בן-גוריון)
יום שלישי, 24.11.2009, 11:30
חדר 1061, בניין מאייר, הפקולטה להנדסת חשמל

Full reference image Quality Index (QI) is an important tool for a cheap and fast assessment of image degradation by different algorithms. A short survey and comparison of recent gray-scale and color image QI is presented. We choose to divide degradation of images into two: luminance changes and saturation changes. From the comparison we see that, in the case of luminance changes, gray scale QI better correlates with Human Visual System (HVS) test results. On the other hand gray scale QI fails to predict saturation changes in color images. A discussion on color saturation and image quality is presented. Based on this discussion and the previous analysis we develop a new QI for color images. The new QI is based on SSIM and Quaternions. We show that Quaternion SSIM (QSSIM) is a single QI which can successfully predict both structural and saturation changes in color images.

בחזרה לאינדקס האירועים