Omnidirectional stereovision system for occupancy grid

Abstract

Vision systems provide large amount of data to robots about its environment. Particularly, an omnidirectional vision system provides information in all directions in just one image. By processing a pair of omnidirectional images it is possible to obtain distances between the robot and the objects in its working environment. Using only an omnidirectional stereovision system as source of information to create a stochastic representation of the environment, known as occupancy grids, a robot can determine the probability of occupation of the space and navigate autonomously. This article shows a stereo algorithm with a matching restricted by geometrical properties of the vision system using directly the omnidirectional image and a model of that sensor to update an occupancy grid