Abstract

The photoelectric detection system belongs to high-end precision instruments, and its detection methods primarily rely on physical mechanisms and structural designs. With the advancement of 5G and AI technologies, research on data-driven photoelectric systems and detection applications has garnered increasing attention. For batch data of photoelectric devices, edge-side algorithms such as deep learning can significantly enhance the perception capabilities of the equipment. Techniques like multimodal approaches, multiphysics coupling, and data augmentation enable the fusion of spatial and temporal data to achieve ultra-low concentration/high-resolution accurate identification in photoelectric detection. Finally, this report demonstrates the accurate identification of photoelectric devices for various types of gases with over 99.6% precision through AI algorithms such as spatiotemporal data stitching, neural network methods, and PCA.

Biography

Jianxiong Zhu is an academic affiliated with Southeast University, China. He is engaged in teaching, research, and scholarly activities, contributing to the university’s strong tradition of academic excellence and innovation. His work reflects a commitment to advancing knowledge, fostering interdisciplinary collaboration, and supporting the development of students and the broader research community in China.