Project Description:
FALCON is a real-time human tracking system that combines mmWave radar and computer vision to maintain continuous identity and 3D position even during complete visual occlusion. Traditional camera-based systems fail when a person is blocked by walls or obstacles, but our system leverages a TI IWR6843 radar alongside an Intel RealSense camera to overcome this limitation. Using a fusion engine with Kalman filtering and sensor handoff logic, FALCON seamlessly transitions between vision and radar in under two frames with zero identity loss. The system runs in real time on an embedded NVIDIA Orin Nano and supports multi-person tracking, making it suitable for applications in robotics, security, and assistive technologies.
