Real-time Occupancy Detection: Modified Background Subtraction with Otsu's Dynamic Thresholding Paradigm

This paper introduces a modified background subtraction algorithm for real-time occupancy detection in both static and dynamic scenarios. The system, validated through rigorous testing with an ESP32CAM-based setup, delivers accurate insights and reduces false positives while accurately identifying true occupancy events, as indicated by an F1 Score of 0.93. The approach offers efficient, real-time occupancy monitoring without complexity, bridging the gap between manual methods and automated solutions. Conventional manual methods are labor-intensive and inefficient, necessitating automated solutions. This research provides a practical, real-time solution for occupancy detection, essential for energy conservation and automation systems