Major orthogonal vector-based local localization algorithm for indoor mobile robots

A low-complexity localization algorithm with high degree of robustness and accuracy is needed for navigation of mobile robots in real-time embedded robotic systems. In this paper we describe an efficient local localization algorithm for mobile robots using major orthogonal vectors that represent maj...

Full description

Saved in:
Bibliographic Details
Main Authors: Sang Woo Choi, Byung Kook Kim
Format: Conference Proceeding
Language:English
Subjects:
Computational complexity
Computational efficiency
Data mining
Feature extraction
Least squares methods
Mobile robots
Navigation
Real time systems
Robustness
Simultaneous localization and mapping
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:A low-complexity localization algorithm with high degree of robustness and accuracy is needed for navigation of mobile robots in real-time embedded robotic systems. In this paper we describe an efficient local localization algorithm for mobile robots using major orthogonal vectors that represent major elements of structured indoor environments. We propose major orthogonal vector extraction algorithm using a novel orthogonal least square fitting method and a major orthogonal vector matching algorithm. We also propose a fast position estimation with solutions for both nonsingular and singular cases. The main contribution is reducing the complexity of the matching process, which is the most complex step in local localization algorithm for mobile robots. We examine validity of our algorithm with various simulations and experiments, and also revealed accuracy and reduction of computational time of the matching process compared to conventional localization algorithms.
ISSN:2163-5137
DOI:10.1109/ISIE.2009.5216055