In the hazy weather, the image of outdoor scene is degraded by suspended particles. Recently, great progresses have been made on single image haze removal. Therefore an accurate quality metric is required. Although image quality assessment has been extensively studied these past few decades, few methods can be applied in this area since there is no reference image. This paper proposed an evaluation metric combining the ascension of contrast degree with the structural similarity to deal with the problem. The ascension of the contrast degree and the structural similarity are computed respectively based on the spatial frequency contrast and the edges consistency between the hazy and the dehazed image. At last, we evaluate the state-of-the-art image haze removal methods based on our evaluation metric.
Dehaze Image Quality Assessment Contrast Enhancement the Contrast Degree of Ascension Structural Similarity
Shuai Fang Jingrong Yang Jiqing Zhan Hongwu Yuan Ruizhong Rao
School of Computer and Information, Hefei University of Technology, Hefei 230009, China and Anhui In School of Computer and Information, Hefei University of Technology, Hefei 230009, China School of Computer and Information,Hefei University of Technology, Hefei 230009, China Anhui Institute of Physical Science, Chinese Academy of Sciences, Hefei 230031,China Anhui Institute of Optics and Fine Mechanics,Chinese Academy of Sciences, Hefei 230031, China