Abstract: To solve the problem of visual imaging degradation caused by the environmental factors of high dust and fog and low illumination in coal mine comprehensive excavation working face, the enhancement method for coal mine degraded images is proposed. Firstly, aiming at the problem of environmental illumination estimation in the dehazing process, the dust and fog distribution function of the degraded image is constructed according to the RGB channel difference and the lightness-saturation difference of the degraded image, and the dust and fog region is segmented by the maximum class variance method. Guided by the dust and fog segmentation map, the gradient domain guided filtering is used to realize the adaptive pixel-by-pixel estimation of the environmental illumination of the degraded image. Secondly, considering the glow effect defect of the degraded image, the glow term is obtained by introducing the near-light source coefficient and combining with the environmental light distribution map, so as to correct the traditional atmospheric scattering model by glow removal. Furthermore, in order to solve the problem of accurate transmission estimation in the dehazing process, the bright channel prior is introduced to supplement the original dark channel prior, so as to realize the smooth transition and adaptive processing of the dark and bright areas of the degraded image. The brightness perception weight combined with the Gaussian weight dynamic allocation function is designed to perform weighted fusion of dark and bright dual-channel transmission. The corrected atmospheric scattering model is used to realize the inversion dehazing of the degraded image. Then, in order to solve the problem of local over-exposure and color distortion in the process of brightness correction, based on the complementary characteristics of arc-tangent transformation enhancement algorithm and self-calibrated illumination learning algorithm, the optimized components of the two algorithms are extracted by adaptive gamma correction function, and the pixel-by-pixel brightness correction of the degraded image is realized by weighted fusion of the optimized components. Finally, the dust and fog removal and brightness correction experiments are carried out. The experimental results showed that, based on the qualitative and quantitative analysis of the dust and fog removal experiments, the proposed method has the best dehazing effect compared to algorithms such as DehazeNet, MSCNN, and CEEF, and there is no structural distortion in the dehazing images; the average dust and fog removal rates of DCP, CAP, DehazeNet, MSCNN, and CEEF algorithms for 50 sets of dust and fog images are 49.9%, 63.2%, 62.2%, 47.8%, and 74.5%, respectively, while the average dust and fog removal rate of the proposed method is as high as 83.5%. Based on the qualitative and quantitative analysis of brightness correction experiments, it can be seen that compared with algorithms such as MSR, URetinex Net, Retinex Net, etc., the proposed method has the highest color fidelity for brightness corrected images and there is no local overexposure problem; the mean chroma error of MSR, URetinex Net, Retinex Net, Zero DCE, and IceNet algorithms for 50 sets of low light images are 55.9, 54.8, 53.0, 32.3, and 14.3, respectively, while the mean chroma error of the proposed method is only 7.7. By analyzing the complex environmental characteristics of coal mine comprehensive excavation working face, the method realizes the enhancement of coal mine complex environmental degradation images, so as to provide technical support for the intelligent perception of coal mine personnel-equipment-environment.