A novel video dehazing method based on temporal visual coherence
Xinguang Xiang, Yang J. Cheng, Jinhui Tang · 2015
In this paper, we propose a novel video dehazing scheme which mainly consists of atmospheric light estimation, transmission map extraction and optimization by utilizing motion estimation. The key contribution of our scheme is the video temporal visual coherence improvement after video dehazing. We first extract the transmission map frame-by-frame by using dark channel prior model and then estimate motion vectors in original video without dehazing. The motion vectors are used to improve the temporal visual coherence on the transmission map sequence. To avoid blocking-artifacts, guided filter can also be used to refine the transmission map. Experimental results show that our model can achieve a good dehazing effect.