Laser-Guided Control for Multicopter UAVs: Preliminary Design and Testing

This paper presents a preliminary study of a laserguided, vision-based framework for evaluating and tuning multicopter trajectory tracking. A floor line and a UAV-projected laser dot are captured by a downward-facing onboard camera, and an offline image-processing pipeline computes the signed lateral deviation between them for proportional-integral-derivative (PID) tuning. We compare the vision-derived error directly against OptiTrack on straight X/Y trajectories at fixed altitude. Experiments show that the vision metric gives more repeatable and less variable lateral-error signals than OptiTrack, with mid-range proportional and derivative gains yielding the most stable tracking, while excessive integral action degrades performance.Overall, the method delivers a practical, vision-based means for controller tuning and lays the groundwork for future real-time visual control of UAVs.