The parallel simulation technique of scanning electron microscope imaging based on Monte Carlo, electron tracking algorithm and geometric space construction is of great value in studying the properties of complex-shaped solids and semiconductor detection, which helps to improve the design of scanning electron microscope and understand the material material and structure corresponding to its imaging, and helps to improve the accuracy of semiconductor detection. Li Huimin started the numerical simulation of scanning electron microscope imaging with complex aggregate structure around 2003, and realized the massively parallelization, etc., which laid the foundation for the related research afterwards. This paper reviews the background and history of this research and the subsequent expansion of semiconductor linewidth detection based on this research.