To completely remove the haze effect while avoiding shortcomings,such as halo artifacts and low brightness,this paper proposes an image dehazing method based on adaptive dark channel prior and median atmospheric light. In this method,an adaptive dark channel prior is introduced,which adaptively determines the neighborhood size for calculating the dark channel based on the brightness,saturation,and standard deviation of each pixel′s neighborhood,thereby estimating the transmittance dynamically to maintain appropriate image brightness and avoid halo artifacts.The top 10% brightest pixels are selected,and the median value of these pixels in each of the three channels is taken as the atmospheric light value.Based on the estimated transmittance and atmospheric light,the hazed image is restored using the atmospheric scattering model.Experimental results demonstrate that the proposed method outperforms existing methods in terms of visual effect,with information entropy and average gradient improved by 2.12% and 5.58%,respectively,and a more reasonable histogram distribution. Therefore, the proposed method can be more effectively applied to image dehazing.