Real-Time Image Processing in Autonomous UAVs for Enhanced Rescue Operations

Unmanned Aerial Vehicles (UAVs) are increasingly utilized in rescue operations, offering the ability to rapidly assess and respond to critical situations. This paper presents an autonomous UAV system designed for enhanced rescue operations through real-time image processing. By employing the YOLOv8n deep learning model for object detection, the UAV is capable of accurately identifying survivors, obstacles, and critical objects in disaster zones with high speed and efficiency. The system architecture includes a high-resolution FPV camera for data capture, a video transmitter for seamless data transmission to ground stations, and a processing pipeline optimized for immediate object detection and decision-making. Real-time data processing enables rapid identification of targets and alerts rescue teams through integrated ground communication. This autonomous UAV solution demonstrates significant advantages over traditional survey methods, such as reduced latency, improved accuracy, and enhanced flexibility in rugged environments, making it a valuable tool for modern rescue operations. The proposed system not only improves response time but also bolsters the safety and effectiveness of rescue missions.