Image Processing Technology in Fog and Haze Environment Based on Retinex Theory

Abstract

For the multi-scale Retinex algorithm (MSR), the image contrast and image information is not ideal enough. In this paper, we propose an algorithm of image haze removal based on the fusion of dark prior and Retinex theory. Firstly, the algorithm of dark primary color is used to restore the haze image. Then, the multi-scale Retinex algorithm is applied to enhance the haze. Finally, the average value of brightness, standard deviation, entropy, mean square error, and peak signal-to-noise ratio were used as evaluation criteria for image enhancement. The simulation results of Matlab software show that the smog image processed by this algorithm has increased image contrast and more image information. We conduct experiments to prove that the proposed algorithm can provide a better representation.