Haze Level Evaluation Using Dark and Bright Channel Prior Information

Haze level evaluation is highly desired in outdoor scene monitoring applications. However, there are relatively few approaches available this area. In paper, a novel haze strategy for real-world scenes presented. The idea inspired by the utilization of dark and bright channel prior (DBCP) removal. change between hazy haze-free channels could serve as indicator, we have named it DBCP-I. variation contrast single image also contains useful information to reflect level. By searching segmentation threshold, metric called DBCP-II proposed. Combining strengths above two indicators, hybrid DBCP-III constructed achieve better performance. experiment results on public, benchmark datasets show advantages proposed methods terms assessment accuracy with subjective human ratings. study first-of-its-kind preliminary exploration field real scenes, has great potential promote research autonomous driving automatic air quality monitoring. open-source codes algorithms free download.