Global Localization Using Low-frequency Image-based Descriptor and Range Data-based Validation

Chansoo Park, Jae-Bok Song

Research output: Contribution to journalArticle

Abstract

In image-based global localization, a robot pose is estimated through image association when the robot revisits a previously visited location on a map. Image association is typically performed using high-level local features such as scale invariant feature transform (SIFT) and speeded up robust feature (SURF). However, these methods suffer from false-positive association and high computational load to reject outliers. In this study, we introduce a novel global localization method based on the proposed low-frequency image-based descriptor (LFID) and laser range data. The image is first processed by reducing the range of luminance in the frequency domain. Visual features are then extracted from the processed image through a kernel window. These visual features are described as binary representation for fast association. Because this binary representation includes a spatial distribution of features, it can minimize false-positive association. Nevertheless, false-positive association could occur when scenes appear to be similar from different viewpoints. To address this problem, this study adopts a laser rangefinder to validate the similarity of the place and reject false-positives from the scene recognition. Experimental results confirm the effectiveness of the proposed scheme in actual indoor environments.

Original languageEnglish
Pages (from-to)1-9
Number of pages9
JournalInternational Journal of Control, Automation and Systems
DOIs
Publication statusAccepted/In press - 2018 May 15

Fingerprint

Robots
Range finders
Lasers
Spatial distribution
Luminance

Keywords

  • Global localization
  • low-frequency image-based descriptor
  • range data validation

ASJC Scopus subject areas

  • Control and Systems Engineering
  • Computer Science Applications

Cite this

@article{08cfad8b95d84582ae2887dd8b1d4fed,
title = "Global Localization Using Low-frequency Image-based Descriptor and Range Data-based Validation",
abstract = "In image-based global localization, a robot pose is estimated through image association when the robot revisits a previously visited location on a map. Image association is typically performed using high-level local features such as scale invariant feature transform (SIFT) and speeded up robust feature (SURF). However, these methods suffer from false-positive association and high computational load to reject outliers. In this study, we introduce a novel global localization method based on the proposed low-frequency image-based descriptor (LFID) and laser range data. The image is first processed by reducing the range of luminance in the frequency domain. Visual features are then extracted from the processed image through a kernel window. These visual features are described as binary representation for fast association. Because this binary representation includes a spatial distribution of features, it can minimize false-positive association. Nevertheless, false-positive association could occur when scenes appear to be similar from different viewpoints. To address this problem, this study adopts a laser rangefinder to validate the similarity of the place and reject false-positives from the scene recognition. Experimental results confirm the effectiveness of the proposed scheme in actual indoor environments.",
keywords = "Global localization, low-frequency image-based descriptor, range data validation",
author = "Chansoo Park and Jae-Bok Song",
year = "2018",
month = "5",
day = "15",
doi = "10.1007/s12555-016-0491-y",
language = "English",
pages = "1--9",
journal = "International Journal of Control, Automation and Systems",
issn = "1598-6446",
publisher = "Institute of Control, Robotics and Systems",

}

TY - JOUR

T1 - Global Localization Using Low-frequency Image-based Descriptor and Range Data-based Validation

AU - Park, Chansoo

AU - Song, Jae-Bok

PY - 2018/5/15

Y1 - 2018/5/15

N2 - In image-based global localization, a robot pose is estimated through image association when the robot revisits a previously visited location on a map. Image association is typically performed using high-level local features such as scale invariant feature transform (SIFT) and speeded up robust feature (SURF). However, these methods suffer from false-positive association and high computational load to reject outliers. In this study, we introduce a novel global localization method based on the proposed low-frequency image-based descriptor (LFID) and laser range data. The image is first processed by reducing the range of luminance in the frequency domain. Visual features are then extracted from the processed image through a kernel window. These visual features are described as binary representation for fast association. Because this binary representation includes a spatial distribution of features, it can minimize false-positive association. Nevertheless, false-positive association could occur when scenes appear to be similar from different viewpoints. To address this problem, this study adopts a laser rangefinder to validate the similarity of the place and reject false-positives from the scene recognition. Experimental results confirm the effectiveness of the proposed scheme in actual indoor environments.

AB - In image-based global localization, a robot pose is estimated through image association when the robot revisits a previously visited location on a map. Image association is typically performed using high-level local features such as scale invariant feature transform (SIFT) and speeded up robust feature (SURF). However, these methods suffer from false-positive association and high computational load to reject outliers. In this study, we introduce a novel global localization method based on the proposed low-frequency image-based descriptor (LFID) and laser range data. The image is first processed by reducing the range of luminance in the frequency domain. Visual features are then extracted from the processed image through a kernel window. These visual features are described as binary representation for fast association. Because this binary representation includes a spatial distribution of features, it can minimize false-positive association. Nevertheless, false-positive association could occur when scenes appear to be similar from different viewpoints. To address this problem, this study adopts a laser rangefinder to validate the similarity of the place and reject false-positives from the scene recognition. Experimental results confirm the effectiveness of the proposed scheme in actual indoor environments.

KW - Global localization

KW - low-frequency image-based descriptor

KW - range data validation

UR - http://www.scopus.com/inward/record.url?scp=85046893169&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85046893169&partnerID=8YFLogxK

U2 - 10.1007/s12555-016-0491-y

DO - 10.1007/s12555-016-0491-y

M3 - Article

AN - SCOPUS:85046893169

SP - 1

EP - 9

JO - International Journal of Control, Automation and Systems

JF - International Journal of Control, Automation and Systems

SN - 1598-6446

ER -