0
research-article

Appearance-based localization of mobile robots using Group LASSO regression

[+] Author and Article Information
Huan Do

Research Associate, School of Computer Science, University of Adelaide, South Australia, 5005
huan.do@adelaide.edu.au

Jongeun Choi

Associate Professor, School of Mechanical Engineering, Yonsei University, Seoul, South Korea
jongeunchoi@yonsei.ac.kr

Chae Young Lim

Associate Professor, Department of Statistics, Seoul National University, Seoul, South Korea, 08826
limc@stats.snu.ac.kr

Tapabrata Maiti

Professor, Department of Statistics and Probability Michigan State University, East Lansing, MI 48824, U.S.A
maiti@stt.msu.edu

1Corresponding author.

ASME doi:10.1115/1.4039286 History: Received June 10, 2017; Revised January 20, 2018

Abstract

Appearance-based localization is a type of robot self-navigation technique, in which the environment map describes a visual features map instead of a geometrical features map. Since images are high dimensional, commonly learning schemes are developed based on features space instead of image space. Therefore, the localization performance essentially depends on the set of chosen visual features. For a high dimensional feature space, choosing the optimal set of features by handcrafting is impractical. Thus, we build a regression model based on extracted visual features from raw images as predictors to estimate the robot's location in 2-D coordinates. We define our supervised learning problem as: given the training data, our model finds the optimal subset of the features that maximizes the localization performance. To tackle the problem, we propose an integrated localization model that consists of two main components: the Least Absolute Shrinkage and Selection Operator (LASSO) regression followed by a filtering estimator. In this study, we examine two candidates for the filtering estimator: the extended Kalman filter (EKF) and Particle Filter (PF). From a raw image, we extract a number of visual features, viz. Fast Fourier Transform, color histogram, and the Speeded-Up Robust Features (SURF). Our method is implemented in both indoor mobile robot and outdoor vehicle equipped with an omni-directional camera. The results validate the effectiveness of our proposed approach.

Copyright (c) 2018 by ASME
Your Session has timed out. Please sign back in to continue.

References

Figures

Tables

Errata

Discussions

Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging and repositioning the boxes below.

Related Journal Articles
Related eBook Content
Topic Collections

Sorry! You do not have access to this content. For assistance or to subscribe, please contact us:

  • TELEPHONE: 1-800-843-2763 (Toll-free in the USA)
  • EMAIL: asmedigitalcollection@asme.org
Sign In