Efficient image dehazing using multi-objective differential evolution
Sukhdeep Kaur, Navleen Kaur · 2017
Digital image haze removal algorithms become more valuable for most vision applications. It is found that majority of researchers has ignored lots of challenges such as the problem of halo artifacts, edge preservation, color distortion, etc. To get rid of the issues with previous research, the latest integrated algorithm is proposed. The proposed technique has modified well-known dark channel prior based dehazing technique by using the multi-objective optimization fitness function. Adaptive histogram equalization has also been used to remove the saturated pixel issue of the digital image haze removal. The proposed technique is tested on five well-known hazy images. Extensive analysis has shown that the proposed method can remove the limitations of existing methods.