Abstract
An Educational Tool to Support Learning Robot VisionWith the rising popularity of robotics in our modern world there is an increase in engineeringprograms that offer an introductory course in robotics. This common introductory roboticscourse generally covers the fundamental theory of robotics including robot kinematics,dynamics, differential movements, trajectory planning and basic computer vision algorithmscommonly used in the field of robotics. To teach robotic vision the student is generallyexposed to a variety of vision algorithms where they learn how to combine them along withthe selection of their parameters to produce a functional vision system.This paper introduces an educational software tool that allows the student to create theircomplete robotic vision system. The tool consists of an integrated development environment(IDE) where the student practices with the different methods and parameters and learns whatcombination works best for their purpose. While this may appear like a common imageprocessing tool, it’s quite different in that the tool only performs the basic methods studied inthe course. This includes image histogram plots, image threshholding, low and high pass FastFourier Transform (FFT) based filters, 3x3 mask convolution, Hit-Miss morphologicaltransform, Hough transforms to detect Lines, Hough transforms to detect circles, randomnoise addition, image moment computations, and image negative generation. The engineeringof the vision system is not performed by the tool, it is only assisting in the process. Becausethe tool is an open system, once the student decides on the best combination of algorithmsand their corresponding parameters, the student can then implement a complete vision systemby accessing the same library of methods used to implement the tool thereby saving themimplementation time.
