Single image haze removal on complex imaging background

A haze removal method for single image on complex imaging background is proposed. The basic assumption of dark channel prior and the selecting principle of atmospheric light value are studied theoretically, and it is found that the template scale plays a key role in the determination of atmospheric light value and the quality of transmission image, which are two crucial parameters for getting high-precision images. Based on this, a multi-scale model is ingeniously proposed in order to obtain the most desirable two crucial parameters simultaneously. The advantages of the multi-scale model for single image haze removal on complex imaging background are also experimentally demonstrated.