Outdoor view recognition based on landmark grouping and logistic regression

Vision-based robot localization outdoors has remained more elusive than its indoors counterpart. Drastic illumination changes and the scarceness of suitable landmarks are the main difficulties. This paper attempts to surmount them by deviating from the main trend of using local features. Instead, a...

Full description

Bibliographic Details
Authors: Todt, Eduardo, Torras, Carme
Format: article
Status:Versión aceptada para publicación
Publication Date:2013
Country:España
Institution:Consejo Superior de Investigaciones Científicas (CSIC)
Repository:DIGITAL.CSIC. Repositorio Institucional del CSIC
OAI Identifier:oai:digital.csic.es:10261/96438
Online Access:http://hdl.handle.net/10261/96438
Access Level:Open access
Keyword:Visual landmarks
Autonomous robots
Robot navigation
Visual saliency
id ES_01fd18a67a9fe2cbbc4d0223859f4da3
oai_identifier_str oai:digital.csic.es:10261/96438
network_acronym_str ES
network_name_str España
repository_id_str
spelling Outdoor view recognition based on landmark grouping and logistic regressionTodt, EduardoTorras, CarmeVisual landmarksAutonomous robotsRobot navigationVisual saliencyVision-based robot localization outdoors has remained more elusive than its indoors counterpart. Drastic illumination changes and the scarceness of suitable landmarks are the main difficulties. This paper attempts to surmount them by deviating from the main trend of using local features. Instead, a global descriptor called landmark-view is defined, which aggregates the most visually-salient landmarks present in each scene. Thus, landmark co-occurrence and spatial and saliency relationships between them are added to the single landmark characterization, based on saliency and color distribution. A suitable framework to compare landmark-views is developed, and it is shown how this remarkably enhances the recognition performance, compared against single landmark recognition. A view-matching model is constructed using logistic regression. Experimentation using 45 views, acquired outdoors, containing 273 landmarks, yielded good recognition results. The overall percentage of correct view classification obtained was 80.6%, indicating the adequacy of the approach. © 2013 World Scientific Publishing Company.This work was partially funded by the GARNICS (Gardening with a Cognitive System) project FP7-ICT-247947Peer ReviewedWorld Scientific Publishing2014201420132014info:eu-repo/semantics/articlehttp://purl.org/coar/resource_type/c_6501Postprintinfo:eu-repo/semantics/acceptedVersionhttp://hdl.handle.net/10261/96438reponame:DIGITAL.CSIC. Repositorio Institucional del CSICinstname:Consejo Superior de Investigaciones Científicas (CSIC)Inglés#PLACEHOLDER_PARENT_METADATA_VALUE#info:eu-repo/grantAgreement/EC/FP7/247947http://dx.doi.org/10.1142/S0218001413550045info:eu-repo/semantics/openAccessoai:digital.csic.es:10261/964382026-05-22T06:33:51Z
dc.title.none.fl_str_mv Outdoor view recognition based on landmark grouping and logistic regression
title Outdoor view recognition based on landmark grouping and logistic regression
spellingShingle Outdoor view recognition based on landmark grouping and logistic regression
Todt, Eduardo
Visual landmarks
Autonomous robots
Robot navigation
Visual saliency
title_short Outdoor view recognition based on landmark grouping and logistic regression
title_full Outdoor view recognition based on landmark grouping and logistic regression
title_fullStr Outdoor view recognition based on landmark grouping and logistic regression
title_full_unstemmed Outdoor view recognition based on landmark grouping and logistic regression
title_sort Outdoor view recognition based on landmark grouping and logistic regression
dc.creator.none.fl_str_mv Todt, Eduardo
Torras, Carme
author Todt, Eduardo
author_facet Todt, Eduardo
Torras, Carme
author_role author
author2 Torras, Carme
author2_role author
dc.subject.none.fl_str_mv Visual landmarks
Autonomous robots
Robot navigation
Visual saliency
topic Visual landmarks
Autonomous robots
Robot navigation
Visual saliency
description Vision-based robot localization outdoors has remained more elusive than its indoors counterpart. Drastic illumination changes and the scarceness of suitable landmarks are the main difficulties. This paper attempts to surmount them by deviating from the main trend of using local features. Instead, a global descriptor called landmark-view is defined, which aggregates the most visually-salient landmarks present in each scene. Thus, landmark co-occurrence and spatial and saliency relationships between them are added to the single landmark characterization, based on saliency and color distribution. A suitable framework to compare landmark-views is developed, and it is shown how this remarkably enhances the recognition performance, compared against single landmark recognition. A view-matching model is constructed using logistic regression. Experimentation using 45 views, acquired outdoors, containing 273 landmarks, yielded good recognition results. The overall percentage of correct view classification obtained was 80.6%, indicating the adequacy of the approach. © 2013 World Scientific Publishing Company.
publishDate 2013
dc.date.none.fl_str_mv 2013
2014
2014
2014
dc.type.none.fl_str_mv info:eu-repo/semantics/article
http://purl.org/coar/resource_type/c_6501
Postprint
info:eu-repo/semantics/acceptedVersion
format article
status_str acceptedVersion
dc.identifier.none.fl_str_mv http://hdl.handle.net/10261/96438
url http://hdl.handle.net/10261/96438
dc.language.none.fl_str_mv Inglés
language_invalid_str_mv Inglés
dc.relation.none.fl_str_mv #PLACEHOLDER_PARENT_METADATA_VALUE#
info:eu-repo/grantAgreement/EC/FP7/247947
http://dx.doi.org/10.1142/S0218001413550045
dc.rights.none.fl_str_mv info:eu-repo/semantics/openAccess
eu_rights_str_mv openAccess
dc.publisher.none.fl_str_mv World Scientific Publishing
publisher.none.fl_str_mv World Scientific Publishing
dc.source.none.fl_str_mv reponame:DIGITAL.CSIC. Repositorio Institucional del CSIC
instname:Consejo Superior de Investigaciones Científicas (CSIC)
instname_str Consejo Superior de Investigaciones Científicas (CSIC)
reponame_str DIGITAL.CSIC. Repositorio Institucional del CSIC
collection DIGITAL.CSIC. Repositorio Institucional del CSIC
repository.name.fl_str_mv
repository.mail.fl_str_mv
_version_ 1869402623456575488
score 15,811543

Similar Items