ABSTRACT: Severe flooding is one example of a natural disaster that can result in disastrous losses to property and human life. Unfortunately, the high processing and hardware requirements of these systems make it difficult to implement them in low-resource and low-cost situations. By analyzing all of the data and extracting the water level information on an edge computing device, the quantity of data that needs to be sent to the cloud server is significantly decreased. The enormity and unpredictability of floods are known to result in disastrous losses and disturbances to our everyday existence. Authorities will be able to minimize the disruptions caused by flooding by taking prompt preventive and corrective action if they have access to real-time information on flood risk and conditions.
INTRODUCTION: One of the most important aspects of disaster management is flood detection, which is vital to reducing the damaging effects of floods on infrastructure and communities. Floods are becoming more frequent and more intense globally as climate change continues to intensify extreme weather events like heavy rainfall and rising sea levels. As a result, there is a growing demand for reliable flood detection systems that can quickly and precisely identify flood events, allowing for rapid action and efficient resource allocation. Conventional flood detection techniques frequently depended on human observation and reporting, which can be inaccurate and slow, especially in rural or heavily populated areas. But flood detection skills have been completely transformed by technological developments, especially in the areas of remote sensing, geospatial analysis, and machine learning. With the use of these technologies, flood monitoring procedures might be improved and automated, giving decision-makers access to up-to-date information and useful insights. Through the constant monitoring of wide geographic areas made possible by remote sensing techniques like satellite imaging and radar systems, flood dynamics and extents can be detected with previously unheard-of accuracy and efficiency. By combining many data sources, such as topographic maps, hydrological models, and historical flood records, geospatial analysis improves flood detection even further by identifying vulnerable locations and estimating flood risk. This study is to investigate the efficacy of machine learning, geospatial analysis, and integrated remote sensing approaches for flood detection in this environment. We want to create a complete flood detection framework that can properly identify flood events in real-time, evaluate their repercussions, and expedite reaction and recovery activities by utilizing the synergies between these technologies. Our goal is to show how our strategy may improve flood preparedness and resilience through empirical analysis and validation, which will ultimately help manage flood risk in susceptible areas in a sustainable manner.