Abstract：In underground mining environment, surveillance images are subject to fogging degradation because of dust scattering and it is difficult to conduct environmentally controlled experiments in situ for fogging and defogging investigation. Therefore, this paper defines the degree of image fogging by using the relative average gradient, analyzes the composition of the dust particle sizes in the mine, proposes the method for compose experiment powder to simulate different mining dust, and designs and builds a simulation air chamber to simulate the mining scattering environment. The results indicate that: 1. The degree of fogging defined by the average gradient is better than other commonly used evaluation parameters and its evaluation results are consistent with the human eye perception. The degree of fogging of underground mining images can reach 80%, when the image details merge in the foggy blurring effect. 2. With microscopic observation method we found that 99% of underground mining dust has particle size less than 30 and the proportion with the particle size less than 10 is about 93%. 3. The foggy image generated from the simulated air chamber not only covers the range of the image fogging degree of real mining surveillance images, but also shows the diffusion effect of light scattering, consistent with real surveillance images, that was not produced by the computer-based fogging algorithm, proving the advantage of the air chamber method. This article provides an experimental basis for the study of underground image dehazing method.