Autonomous navigation with discrete markers

Abstract

This project is part of a bigger interdisciplinary project of an autonomous vehicle tour guide that shows visitors around the OSU campus. This project demonstrates the possibility of using discrete markers in autonomous navigation. A smaller scale 1/8 model car was used as the vehicle to demonstrate the viability of the navigation. The car was to navigate a closed square track that was laid out with special markers (ArUco markers) which were separated by at least three feet. The image processing was achieved using OpenCV along with a marker detector library, ArUco, and the motors and navigation behavior were controlled using Python scripting. The end result was a working model car that was able to navigate a closed track laid out with ArUco markers.