Flood Survivor Detection

This project guides a stepwise walkthrough of the research undertaken to explore various Deep Learning and Machine Learning Architectures in-order to solve the problem of Flood-Survivor Detection on LWIR imagery taken by autonomous drones. Survivor detection and localization has been an integral part of search and rescue missions immediately following disasters such as floods or earthquakes. Autonomous Unmanned Aerial Vehicles (UAV) in combination with image processing and machine learning, provide the necessitating tool for the given task at hand. In addition to this, with the growing influence of Machine Learning for a variety of applications and the powerful models which it builds, Object detection is a task which has often been performed. However, it has not been widely used in the context of survivor detection and localization.