Image Dehazing Algorithm By Atmospheric Light Fusion Based on The Prior of Dual-Channel

In order to solve the problems existing in the traditional DCP dehazing algorithm and improve the quality of the image after dehazing, this paper proposes a image dehazing algorithm by atmospheric light fusion based on the prior of Dual-Channel. Firstly, the corresponding atmospheric light and transmittance map are derived from the D channel and bright channel prior models, respectively. Then, the bright channel atmospheric light values obtained by different filtering sizes are fused by wavelet transform. Finally, the linear fusion function is applied to fit the dual-channel atmospheric light estimate, while the dehazed image is calculated using the atmospheric scattering model. Experimental findings indicate that the proposed algorithm is capable of efficiently removing the haze from the original image, and solve the problems of over-processing of the highlight part and low image brightness in the traditional DCP algorithm.