Object Tracker Drone using Computer Vision Techniques under GPS-Jammed Scenarios

FlyHi designs and implements an autonomous quadrotor system that detects and tracks moving objects in GPS-denied environments using computer vision techniques, developed for EHSİM. The system integrates YOLO-based object detection with optical flow and visual odometry for self-localization without GPS. All processing runs in real-time on a compact onboard computer. The approach is validated through both simulation and outdoor flight testing, achieving robust target following performance even under GPS jamming conditions.