Sample Implementations of Vision-Based Mobile Robot Algorithms

Abstract

This chapter presents a detailed, step-by-step demonstration of how vision-based navigation modules can be actually implemented in real life, under 32-bit Windows environment. These lessons start with a simple development of capturing image frames from a running video and then gradually proceeds to more complex tasks of incorporating image processing capabilities e.g. filtering techniques, contrast enhancement, adaptive thresholding etc. Then the lessons demonstrate how to extract path for the robot from such images and how a rulebased approach can be utilized to determine left and right wheel speed settings of a differential drive system.