Aerial imaging systems that offer a birds-eye view of flood-affected locations are incredibly versatile for emergency area analysis. During disasters, many countries rely on UAV (Unmanned Aerial Vehicle) swarms as their first responders to conduct scene analysis at disaster sites. To automate this process, deep learning-based object detection models come into play. These models can be trained to identify and analyze various elements in the scene, providing valuable insights for emergency responders. In the case of flooding, such models can perform real-time analysis of flooded areas, detecting and tracking objects, assessing the extent of damage, and aiding in rescue and recovery operations. This project aims to develop a sophisticated model that can efficiently analyze flooded areas, enabling quick decision-making and effective deployment of resources in emergency situations.
