AquaRescue: Intelligent Drowning Detection and Swift Rescue System Using Integrated Approach

Drowning accidents, in pools continue to be a concern calling for creative solutions to improve safety standards. The research aims to establish a framework for identifying and saving drowning cases in pools. Cutting-edge technologies such as computer vision, Artificial Intelligence, and IoT are integrated into the system to quickly and precisely recognize incidents and trigger rescue operations. Utilizing cameras and advanced algorithms like YOLO v3 and CNN, the system continuously observes the pool environment. By assessing people's movements and locations in the water, it accurately detects drowning cases. Once an incident is identified, the system automatically starts rescue procedures by dispatching rescue tools and notifying lifeguards with alarms across the pool premises. Moreover, it enables emergency calls through GSM SIM900A module in case the lifeguard is not present nearby. This innovative detection and rescue system not only enhances swimmer safety but also reduces the workload on lifeguards, allowing them to react more effectively in emergencies. The ultimate goal of this system, driven by technological advancements, is to minimize drowning incidents in pools. The system's performance has been evaluated against other models like Faster R-CNN, YOLO v3, and SSD. YOLO v3 demonstrated superior accuracy of 90% with combined datasets (GF-1 & GF -2).