Highly monodispersed micron powders are considered to be one of the most promising materials due to their excellent physicochemical properties. However, the current particle classification techniques are restricted by the issues of low-precision and low-efficiency, resulting in uneven particle size distribution and large individual performance differences, making it difficult to meet the application requirements of high-quality products. In this study, aiming at the high-precision cyclone classification of micron powders, a visual experiment platform for particle classification was constructed by combining with 3D printing technology. The influence laws of operation parameters, structure parameters of the cyclone, and physical properties parameters of particles on the classification precision and efficiency were systematically studied. The research results indicated that by optimizing the operating conditions, the classification precision of the target particles can be as high as 95.94%, and the classification efficiency can reach 72.89%. In addition, the mechanism of particle cyclone classification was revealed. The "coarse particles entrainment" in the overflow and "fine particles entrainment" in the underflow phenomena were analyzed, and corresponding solutions were proposed. The results of this study can provide theoretical basis and technical support for the high-precision and high-efficiency particle classification of micron powders.