A vision-based autonomous vehicle tracking robot platform

Abstract

A vision-based autonomous vehicle tracking robot platform has been designed and developed where a video camera mounted in front and on top of the robot is used to capture the environment to provide real-time monitoring and tracking. The image processing is done in MATLAB where the output decision is used to guide the robot's navigation. A differential drive mechanism implemented using two DC motors controlled by a Basic Stamp microcontroller is used. In the field test, the robot has been designed to follow a leading vehicle by tracking a green-colour square placed behind that vehicle. The robot moves forward, reverses, turns left or right autonomously based on the images captured through the camera.