An Improved Probability Density Function for Representing Landmark Positions in Bearing-only SLAM Systems

Huang, Henry and Maire, Frederic D. and Keeratipranon, Narongdech (2007) An Improved Probability Density Function for Representing Landmark Positions in Bearing-only SLAM Systems. In Proceedings 20th Australian Joint Conference on Artificial Intelligence (AI 2007), Gold Coast, Queensland, Australia.

Full text available as: PDF (Author version) - Requires Adobe Acrobat Reader or other PDF viewer.

Abstract

To navigate successfully, a mobile robot must be able to estimate the spatial relationships of the objects of interest in its environment accurately. The main advantage of a bearing-only SLAM (Simultaneous Localization and Mapping) system is that it requires only a cheap vision sensor to enable a mobile robot to gain knowledge of its environment and navigate. In this paper, we focus on the representation of the spatial uncertainty of landmarks caused by sensor noise. This representation is critical for the initialization of landmark positions in bearing-only SLAM systems. We follow a principled approach for computing the Probability Density Functions (PDFs) of landmark positions when an initial observation is made. We characterize the PDF p(r,\alpha) of a landmark position expressed in polar coordinates when r and \alpha are independent, and the marginal probability p(r) of the PDF is constrained to be uniform. Existing methods that approximate the PDF of a landmark position with a mixture of Gaussians do not satisfy this uniformity requirement. We also show how to use the proposed PDFs for a 2D bearing-only SLAM system that relies only on the landmark bearings measured by a panoramic camera.

Item Type:	Conference Paper
Status:	In Press
Keywords:	SLAM, Probability Density Function
Subjects:	280000 Information, Computing and Communication Sciences > 280200 Artificial Intelligence and Signal and Image Processing > 280209 Intelligent Robotics 280000 Information, Computing and Communication Sciences > 280200 Artificial Intelligence and Signal and Image Processing 280000 Information, Computing and Communication Sciences
ID Code:	9397
Deposited By:	Huang, Henry
Deposited On:	10 September 2007
Alternative Locations:	http://www.cit.gu.edu.au/conferences/austai/
Copyright Owner:	Copyright 2007 (please consult author)