Home /Research /Inexact navigation
OTHER

Inexact navigation

Karen T. Sutherland, William B. Thompson

Year
2002
Citations
42

Abstract

Robot navigation based on determining the relative position of visible landmarks and then using this information to estimate the robot's location is considered. Accuracy is difficult to obtain, because measurements are approximate and combining them can lead to large errors in localization. However, a careful choice of landmarks can decrease errors before they occur. The problem of determining viewpoint position, given relative visual angles between landmarks, is addressed. It is shown that for any given error in angle measurement the error in localization can vary significantly, depending on the configuration of landmarks used. The authors demonstrate how an algorithm which makes a wise choice of landmark configurations can decrease error. The algorithm's performance is compared to that of using landmarks chosen indiscriminately.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>

Keywords

LandmarkPosition (finance)Computer visionArtificial intelligenceComputer scienceRobotAlgorithm

Related papers

Browse all OTHER papers