Implementing HOG & AMDF based shape detection algorithm for computer vision & robotics education using LEGO Mindstorms NXT

Abstract

Using mobile robots in the engineering education is a well-accepted and practical way to teach in the undergraduate programs. In the literature, several different cases are reported to show the adaptation of some robots. This paper describes how mobile robots and basic image processing algorithms might be combined in the MATLAB programming environment to provide a more efficient learning and simulation platform for engineering freshman students. In this study the RWTH - Mindstorms NXT Toolbox for MATLAB has been used and some module have been extended with basic image processing algorithms to control the LEGO Mindstorms robots via Bluetooth.