Paper Info
Title
Biologically-Inspired Visual Landmark Learning for Mobile Robots
Abstract
This paper presents a biologically-inspired method for selecting visual landmarks which are suitable for navigating within not pre-engineered environments. A landmark is a region of the goal image which is chosen according to its reliability measured through a phase called Turn Back and Look (TBL). This mimics the learning behavior of some social insects. The TBL phase affects the conservativeness of the vector field thus allowing us to compute the visual potential function which drives the navigation to the goal. Furthermore, the conservativeness of the navigation vector field allows us to assess if the learning phase has produced good landmarks. The presence of a potential function means that classical control theory principles based on the Lyapunov functions can be applied to assess the robustness of the navigation strategy. Results of experiments using a Nomad200 mobile robot and a color camera are presented throughout the paper.
Year
DOI
Venue
1999
10.1007/3-540-40044-3_9
EWLR
Keywords
Field
DocType
control theory,vector field,mobile robot,lyapunov function
Computer vision,Lyapunov function,Vector field,Visual landmarks,Robustness (computer science),Artificial intelligence,Classical control theory,Engineering,Landmark,Mobile robot
Conference
Volume
ISSN
ISBN
1812
0302-9743
3-540-41162-3
Citations 
PageRank 
References 
2
1.02
5
Authors
2
Name
Order
Citations
PageRank
Giovanni M. Bianco1185.52
Riccardo Cassinis2214.20