Real-Time Drowning Detection System Using OpenCV and Deep Learning

Drowning is the 3rd leading cause of unintentional injury death worldwide, accounting for 7% of all injuryrelated deaths, with an estimated 320,000 annual fatalities. Safety in swimming pools remains a critical concern. This paper presents a real-time drowning detection system that leverages OpenCV and deep learning to identify drowning incidents in swimming pools and trigger alarms for lifeguards. The system processes live video feeds to detect drowning individuals using YOLO (You Only Look Once) and R-CNNs for object detection, combined with movement analysis to distinguish drowning behavior. Experimental results demonstrate the system’s effectiveness in real-world pool environments, achieving high accuracy in detecting drowning incidents. The proposed solution offers a scalable, automated approach to enhance pool safety and reduce drowning-related fatalities.