Image quality assessment based on geometrical structural orientation
Xuan Fei, Zhihui Wei, Liang Xiao, Tianming Zhan · 2010
Image quality assessment takes an important position in image processing field. The structural similarity (SSIM) index describes the similarity of images more appropriately for the human visual system than the mean square error (MSE). By our studying deeply, it is found that the structural information is not represented very well. So there is some weakness about SSIM, especially for assessing the blurred image. By using the structure tensor to get the local contrast and geometrical structural orientation field estimation, an improved index of image quality assessment based on geometrical structural orientation is presented. The experimental results prove its effectiveness.