Detection of Air Pollution in Urban Areas Using Monitoring Images

Air quality monitoring in polluted environments is of great significance to human health. Traditional methods use various pieces meteorological equipment, which have limited applications complex terrains and high costs. In this paper, a novel idea put forward solve the problem air pollution urban areas. We investigate whether can be assessed visually by examining haziness photos from far distance. Specifically, correlation between indexes, such as AQI, PM2.5, PM10, real outdoor scenarios level evaluation scores images calculated. The results show that objective indicators indeed reflect pollution, degree invariant image size. To apply new observation practical system, method called fastDBCP (fast dark bright channel prior) developed. Based on down-sampling strategy, ratio calculated prior information scaled adopted visual index pollution. Our experimental proposed metric not only demonstrates advantages terms its computational speed, but also shows classification accuracy compared competing metrics.