Vision-Based Real Time Guidance of UAV

This paper presents vision-based estimator and a new guidance method for standoff tracking by unmanned aerial vehicle (UAV). Vision information obtained from a single camera, fixed to the vehicle as well as navigation and communications hardware, UAVs are capable of surveying, identifying and tracking ground, sea or aerial targets. The visual measurements of the target can be transformed into bearing measurements using the vehicle position and orientation, camera orientation with respect to the vehicle, and the pixel location of the target. Lyapunov vector field-based guidance law system is utilized to generate airspeed and heading rate control inputs which guide the UAV to circle about a designated ground target with a desired standoff radius. A vector field guidance law is designed for a stationary target with no wind and then modified with a correction term that accounts for a moving target and constant back-ground wind.