An Intelligent IoT-Driven ATM Security Framework With YOLOv5-Based Object Recognition

In the era of digitalization, advanced technology is employed to enhance the safety and security of ATM users, ensuring the integrity of banking operations. The Real-time Automatic ATM Booth Security System aims to provide an innovative and effective solution by integrating sophisticated monitoring, detection, and reaction capabilities. To mitigate different types of robberies, we propose a security system for ATMs that detects specific threat objects based on real-time image analysis using the YOLOv5 Object Detection Algorithm. This system leverages an Arduino-based embedded platform to process real-time data collected through sensors. The module is designed to recognize and classify potential weapons, including screwdrivers, scissors, hammers, and rods. Upon detecting an individual carrying a weapon, the system generates an instantaneous alarm signal and autonomously triggers the closure of the ATM booth door, preventing the suspect from escaping. Another key feature of the proposed system is the real-time notification to the control center of the nearest police station and the respective bank. This approach ensures enhanced monitoring and control, ultimately improving ATM security.