AQUA-ALERT: A Computer Vision-Based Real-Time Water Level Monitoring System for Underpasses

During the monsoon season, underpasses flood severely, causing traffic jams, auto accidents, and fatalities in many countries. This set of problems regarding waterlogging, especially in the underpasses of any countries, forms the principal set of objectives of the project to implement an accurate, real-time water-level monitoring system. What makes this system unique is its converting standard surveillance cameras into smart environmental sensors through a sophisticated computer vision pipeline thereby providing a new, cost-effective, and scalable alternative to the conventional physical measurement infrastructure. Rather than using costly and fragile, physical sensors, the proposed system utilizes a typical surveillance camera along with image-processing techniques and some calibrated visual scale to infer water depth with adequate precision. Accurate and consistent measurements will be guaranteed by sophisticated image processing and data filtering techniques. The system automatically captures and stores images that match crucial water level conditions, providing a low maintenance, scalable, and cost-effective approach to flood control and public safety and sending them to the Central Control Room (CCR).