Vision Based Alignment to Runway during Approach for Landing of Fixed wing UAVs

Aligning a fixed wing UAV to the runway using vision as the only sensor has been implemented. FlightGear and Matlab is interfaced to develop a Vision-in-the-Loop Simulation facility. Interfacing the image to control in real time has been successfully implemented to autonomously align the UAV to the runway. Image processing is performed to detect and track the runway lines as visual cues using Hough Transforms. An image based proportional controller consisting of course and fine roll control inputs, akin to a manned aircraft pilot input has been developed for UAV alignment. Simulation results with varying heading off-set to the runway have been presented.