Previous studies have revealed that toxic gases and dust (smoke dust) are the most common pollutants generated by the blasting operations in open-pit mines, which might lead to a threat to the environment’s condition, health and safety, and properties protection around the blasting site. In order to deal with the problems, a pollution evaluation system was established based on the fractal dimension theory (Dbox(P)) and grayscale average algorithm (Ga) in digital image-processing technology to recognize and analyze the distributions of the smoke-dust cloud, and subsequently determine the pollution degrees. The computation processes of Dbox(P) and Ga indicate three fitted correlations between the parameters and diffusion time of smoke dust. Then, a pollution index (Pi) is put forward to integrate the global and local features of Dbox(P) and Ga, and develop a hazard classification mechanism for the blasting pollutants. Results obviously denote three diffusion stages of the pollutants, mainly including generation stage, cloud-formation stage, and diffusion stage. In addition, it has been validated that the proposed system can also be utilized in single-point areas within a whole digital image. Besides, there are variation trends of the thresholds T1 and T2 in binarization with the diffusion of pollutants. With this identification and evaluation system, the pollution condition of smoke dust can be obviously determined and analyzed.