标题：A Novel Dehazing Method for Color Fidelity and Contrast Enhancement on Mobile Devices
作者：Wang, Ping; Fan, Qinglan; Zhang, Yunfeng; Bao, Fangxun; Zhang, Caiming
作者机构：[Wang, Ping; Fan, Qinglan; Zhang, Yunfeng] Shandong Univ Finance & Econ, Dept Comp Sci & Technol, Jinan 250014, Shandong, Peoples R China.; [Bao, Fa 更多
通讯作者：Zhang, YF;Zhang, Yunfeng
通讯作者地址：[Zhang, YF]Shandong Univ Finance & Econ, Dept Comp Sci & Technol, Jinan 250014, Shandong, Peoples R China.
来源：IEEE TRANSACTIONS ON CONSUMER ELECTRONICS
关键词：Single image dehazing; region division; local atmospheric light;; iterative algorithm
摘要：Mobile devices that can capture images outdoors often suffer from poor weather conditions. Observed scenes lose contrast and visibility, and are also subject to color shifts due to the presence of fog and haze. In this paper, a single image dehazing method for removing color distortion and enhancing contrast on mobile devices is presented. The method can effectively remove haze from images with an uneven haze density. First, based on the local haze density, a superpixel segmentation algorithm is employed to adaptively divide a hazy image into several regions. Second, the local atmospheric light in each image region is estimated, and an initial transmission map is calculated. Finally, based on the haze imaging model, an iterative algorithm is developed to estimate the local atmospheric light and transmission for restoring the image. The proposed method can effectively recover detailed image information, resulting in an image with natural color, with an acceptable time performance. Experimental results show that the proposed method can be applied effectively to mobile devices, including digital cameras and mobile phones for enhancing the contrast and visibility without introducing color distortion.