Research and Development of a Deep Learning-Based Automated Imaging System for FIB-SEM

With the rapid advancements in semiconductor and materials science, the demand for high-resolution and high-clarity scanning electron microscope (SEM) images continues to grow. To meet this demand, this study proposes an automated artificial intelligence imaging system for FIB-SEM (Focused Ion Beam-Scanning Electron Microscope). The system aims to achieve efficient identification and automatic capture of specific image locations by integrating artificial intelligence-based image recognition and automation control technologies, utilizing a deep learning algorithm (YOLO). This approach enables precise and intelligent imaging of target features. The system is composed of three primary modules: an artificial intelligence system, an automated imaging system, and a human-machine interface (UI), which collectively enable intelligent control and imaging of the FIB-SEM movable stage. Through optimized AI algorithms and automated imaging workflows, the system significantly enhances the efficiency and quality of feature image acquisition.