Vision System for Recognition of Water Level, Rain Water and Flood Detection

Several flood detection techniques have been designed to accurately measure water levels around the environment as a means to mitigate accidents or to be prepared for such scenarios. This paper presents a system that detects water levels and floods within the 2 surveillance cameras. Gathered images were trained into the YOLOv5 algorithm, enabling the system to recognize floods. By operating at a few seconds of intervals, the system can notify users if the water levels are rising around the area. This study concludes that the most optimal position and camera to be used for this study is the outdoor camera, showing better results as seen from the gathered data.