File Download
  Links for fulltext
     (May Require Subscription)
Supplementary

Article: Saliency-Guided Detection of Unknown Objects in RGB-D Indoor Scenes

TitleSaliency-Guided Detection of Unknown Objects in RGB-D Indoor Scenes
Authors
Issue Date2015
PublisherMolecular Diversity Preservation International. The Journal's web site is located at http://www.mdpi.net/sensors
Citation
Sensors, 2015, v. 15 n. 9, p. 21054-21074 How to Cite?
AbstractThis paper studies the problem of detecting unknown objects within indoor environments in an active and natural manner. The visual saliency scheme utilizing both color and depth cues is proposed to arouse the interests of the machine system for detecting unknown objects at salient positions in a 3D scene. The 3D points at the salient positions are selected as seed points for generating object hypotheses using the 3D shape. We perform multi-class labeling on a Markov random field (MRF) over the voxels of the 3D scene, combining cues from object hypotheses and 3D shape. The results from MRF are further refined by merging the labeled objects, which are spatially connected and have high correlation between color histograms. Quantitative and qualitative evaluations on two benchmark RGB-D datasets illustrate the advantages of the proposed method. The experiments of object detection and manipulation performed on a mobile manipulator validate its effectiveness and practicability in robotic applications.
Persistent Identifierhttp://hdl.handle.net/10722/234604
ISSN
2015 Impact Factor: 2.033
2015 SCImago Journal Rankings: 0.546

 

DC FieldValueLanguage
dc.contributor.authorBAO, J-
dc.contributor.authorJIA, Y-
dc.contributor.authorCHENG, Y-
dc.contributor.authorXi, N-
dc.date.accessioned2016-10-14T13:47:59Z-
dc.date.available2016-10-14T13:47:59Z-
dc.date.issued2015-
dc.identifier.citationSensors, 2015, v. 15 n. 9, p. 21054-21074-
dc.identifier.issn1424-8220-
dc.identifier.urihttp://hdl.handle.net/10722/234604-
dc.description.abstractThis paper studies the problem of detecting unknown objects within indoor environments in an active and natural manner. The visual saliency scheme utilizing both color and depth cues is proposed to arouse the interests of the machine system for detecting unknown objects at salient positions in a 3D scene. The 3D points at the salient positions are selected as seed points for generating object hypotheses using the 3D shape. We perform multi-class labeling on a Markov random field (MRF) over the voxels of the 3D scene, combining cues from object hypotheses and 3D shape. The results from MRF are further refined by merging the labeled objects, which are spatially connected and have high correlation between color histograms. Quantitative and qualitative evaluations on two benchmark RGB-D datasets illustrate the advantages of the proposed method. The experiments of object detection and manipulation performed on a mobile manipulator validate its effectiveness and practicability in robotic applications.-
dc.languageeng-
dc.publisherMolecular Diversity Preservation International. The Journal's web site is located at http://www.mdpi.net/sensors-
dc.relation.ispartofSensors-
dc.rightsCreative Commons: Attribution 3.0 Hong Kong License-
dc.titleSaliency-Guided Detection of Unknown Objects in RGB-D Indoor Scenes-
dc.typeArticle-
dc.identifier.emailXi, N: xining@hku.hk-
dc.identifier.authorityXi, N=rp02044-
dc.description.naturepublished_or_final_version-
dc.identifier.doi10.3390/s150921054-
dc.identifier.hkuros269251-
dc.identifier.volume15-
dc.identifier.issue9-
dc.identifier.spage21054-
dc.identifier.epage21074-
dc.publisher.placeSwitzerland-

Export via OAI-PMH Interface in XML Formats


OR


Export to Other Non-XML Formats