Aiming at the problems that traditional dark channel prior dehazing algorithm is prone to produce bright area distortion and the whole dehazing image is dark,an image dehazing algorithm based on dual channel and image quality evaluation model is p roposed. Firstly,the bright region and the non-bright region of the image are divided. Secondly,the atmospheric light value is accurately estimated by using the two-channel prior algorithm. Next,the transmittance of the dark channel prior is taken as the transmittance of the non-bright region,and the transmittance is constructed separately in the brigh t region.The two are fused and refined to obtain the transmittance with parameters.Finally, a parameter-driven image quality evaluation model is constructed to iteratively s elect the optimal fog-free image.The experimental results show that the proposed algorit hm can effectively suppress Halo effect,avoid regional distortion and improve the comp rehensive quality of therestored image.