Drone-based Artificial Intelligence for Efficient Disaster Management: The Significance of Accurate Object Detection and Recognition
S Vedanth, Udit Narayana K B, Sai Harshavardhan, Trupthi Rao, Ashwini Kodipalli
Abstract
In today's rapidly changing world, the significance of accurate and efficient disaster management systems using drones with artificial intelligence cannot be overstated. These advanced systems have the potential to greatly mitigate the effects of natural disasters such as earthquakes, floods, and tsunamis. By incorporating AI-based automatic object recognition capabilities, drones can automatically detect and identify individuals in need of help from above. Once identified, the drones can generate alarms and precisely drop payloads, such as food, clothes, and rescue tools, near the victims. This technology not only improves the efficiency and speed of rescue operations but also reduces the risk for rescue teams by minimizing their exposure to hazardous terrains. Furthermore, the development of this system requires various technologies including drones, cameras, processor boards, and artificial neural networks. By combining these technologies, the system can accurately detect and locate individuals from a height of around 50-100 meters above the ground at a slanted angle. The automated payload dropping mechanism, controlled by the AI system, ensures that essential supplies are delivered to the victims quickly and precisely. In order to accurately detect and identify individuals in need of help, our disaster management system incorporates the advanced YOLO model of version 8. This cutting-edge version of YOLO is specifically designed to efficiently identify and locate people from aerial videos captured by the drone's IR camera. By leveraging the object detection and recognition techniques of YOLO version 8, our system can effectively analyze the video feed in real-time. This allows us to quickly identify and track individuals, even in challenging conditions or downpour.