A Machine Learning-Based Flood Detection and Mobile Alert System for Real-Time Disaster Monitoring Using Remote Sensing Data

Floods are among the most devastating natural disasters, causing severe damage to infrastructure, agriculture, and human lives. Timely identification of flood-affected areas is crucial for effective disaster response. This study proposes a flood image classification approach using Convolutional Neural Networks (CNNs) to detect and classify flood-affected regions in satellite and aerial imagery. The methodology involves data collection, preprocessing, model development, and evaluation. Satellite and aerial images of flood events are gathered and preprocessed to enhance quality and reduce noise. A CNN-based architecture is designed to extract key features, leveraging hierarchical representations of visual data. Transfer learning is applied to fine-tune pre-trained YOLO models on flood datasets, improving efficiency with limited labeled data. To assess performance, experiments are conducted on benchmark datasets using accuracy, precision, recall, and F1-score. The results indicate that YOLO-based flood classification accurately identifies flood-affected areas, performing comparably or better than traditional image processing methods. This system offers valuable insights for disaster response agencies, aiding in resource allocation, evacuation planning, and relief coordination. The proposed YOLO-based flood classification approach provides a reliable and efficient solution for flood detection and monitoring, contributing to improved disaster preparedness and response strategies.