Image quality assessment based on local edge direction histogram
Xiaolin Chen, Rui Zhang, Shibao Zheng · 2011
In this paper, we propose a new edge direction histogram (EDH) based on a structural similarity method for image quality assessment. We were motivated by the fact that the structural information of image can be described by the distribution of gradient magnitudes, as well as the one of edge directions. Therefore, we introduce the edge direction histogram to robustly represent the distribution of edge magnitudes and directions simultaneously. We further develop a distance metric to evaluate the quality of images by computing the structural similarity between EDH descriptors of the reference and the distorted one. Experimental results show that the proposed method outperforms some well-known full-reference image quality metrics.