Journal of Engineering Research

The underground coal mine environment is complex and ever-changing, posing significant challenges for personnel safety behavior management. Traditional monitoring methods suffer from issues such as delayed response and insufficient recognition accuracy. To address these challenges, this study designs an intelligent recognition and early warning system for underground coal mine personnel safety behavior. By integrating machine vision, IoT, and AI technologies, the system enables real-time monitoring, intelligent identification, and proactive alerts for personnel violations. The system establishes a multi-dimensional sensing network to collect operational postures, equipment status, and environmental parameters. Through deep learning models, it performs behavioral pattern analysis and risk assessment, forming a closed-loop management system from data collection to early warning and response. Practical applications demonstrate that this system effectively enhances safety supervision efficiency, reduces accident rates, and provides technical support for underground coal mine production safety.

Underground coal mine; Safety behavior; Intelligent recognition; Early warning system; Deep learning