File Download

There are no files associated with this item.

  Links for fulltext
     (May Require Subscription)
Supplementary

Conference Paper: Multi-body segmentation and motion number estimation via over-segmentation detection

TitleMulti-body segmentation and motion number estimation via over-segmentation detection
Authors
Issue Date2011
PublisherSpringer Verlag. The Journal's web site is located at http://springerlink.com/content/105633/
Citation
Workshop on Computer Vision in Vehicle Technology: From Earth to Mars (CVVT:E2M 2010) in conjunction with the 10th Asian Conference on Computer Vision (ACCV 2010), Queenstown, New Zealand, 8-12 November 2010. In Lecture Notes in Computer Science, 2011, v. 6469 pt. 2, p. 194-203 How to Cite?
AbstractThis paper studies the problem of multi-body segmentation and motion number estimation. It is well known that motion number plays a critical role in the success of multi-body segmentation. Most of the existing methods exploit only motion affinity to segment and determine the number of motions. Motion number estimated in this way is often seriously affected by noise. In this paper, we recast the problem of multi-body segmentation and motion number estimation into an over-segmentation detection problem, and introduce three measures, namely loss of spatial locality (LSL), split ratio (SR) and cluster distance (CD), for over-segmentation detection. A hierarchical clustering method based on motion affinity is applied to split the motion clusters recursively until over-segmentation occurs. Over-segmentation is detected by Kernel Support Vector Machines trained under supervised learning using the above three measures. We leverage on Hopkins155 database to test our method and, with the same motion affinity measure, our method outperforms another state-of-the-art method. To the best of our knowledge, this paper is the first to tackle the problem of multi-body segmentation and motion number estimation from the perspective of over-segmentation detection. © 2011 Springer-Verlag Berlin Heidelberg.
DescriptionLNCS vols. 6468-6469 (pt. 1-2) are the conference proceedings of ACCV 2010
Persistent Identifierhttp://hdl.handle.net/10722/152011
ISSN
2005 Impact Factor: 0.402
2015 SCImago Journal Rankings: 0.252
References

 

DC FieldValueLanguage
dc.contributor.authorPan, Gen_US
dc.contributor.authorWong, Ken_US
dc.date.accessioned2012-06-26T06:32:25Z-
dc.date.available2012-06-26T06:32:25Z-
dc.date.issued2011en_US
dc.identifier.citationWorkshop on Computer Vision in Vehicle Technology: From Earth to Mars (CVVT:E2M 2010) in conjunction with the 10th Asian Conference on Computer Vision (ACCV 2010), Queenstown, New Zealand, 8-12 November 2010. In Lecture Notes in Computer Science, 2011, v. 6469 pt. 2, p. 194-203en_US
dc.identifier.issn0302-9743en_US
dc.identifier.urihttp://hdl.handle.net/10722/152011-
dc.descriptionLNCS vols. 6468-6469 (pt. 1-2) are the conference proceedings of ACCV 2010-
dc.description.abstractThis paper studies the problem of multi-body segmentation and motion number estimation. It is well known that motion number plays a critical role in the success of multi-body segmentation. Most of the existing methods exploit only motion affinity to segment and determine the number of motions. Motion number estimated in this way is often seriously affected by noise. In this paper, we recast the problem of multi-body segmentation and motion number estimation into an over-segmentation detection problem, and introduce three measures, namely loss of spatial locality (LSL), split ratio (SR) and cluster distance (CD), for over-segmentation detection. A hierarchical clustering method based on motion affinity is applied to split the motion clusters recursively until over-segmentation occurs. Over-segmentation is detected by Kernel Support Vector Machines trained under supervised learning using the above three measures. We leverage on Hopkins155 database to test our method and, with the same motion affinity measure, our method outperforms another state-of-the-art method. To the best of our knowledge, this paper is the first to tackle the problem of multi-body segmentation and motion number estimation from the perspective of over-segmentation detection. © 2011 Springer-Verlag Berlin Heidelberg.en_US
dc.languageengen_US
dc.publisherSpringer Verlag. The Journal's web site is located at http://springerlink.com/content/105633/en_US
dc.relation.ispartofLecture Notes in Computer Scienceen_US
dc.titleMulti-body segmentation and motion number estimation via over-segmentation detectionen_US
dc.typeConference_Paperen_US
dc.identifier.emailWong, K:kykwong@cs.hku.hken_US
dc.identifier.authorityWong, K=rp01393en_US
dc.description.naturelink_to_subscribed_fulltexten_US
dc.identifier.doi10.1007/978-3-642-22819-3_20en_US
dc.identifier.scopuseid_2-s2.0-80053119557en_US
dc.identifier.hkuros183620-
dc.relation.referenceshttp://www.scopus.com/mlt/select.url?eid=2-s2.0-80053119557&selection=ref&src=s&origin=recordpageen_US
dc.identifier.volume6469en_US
dc.identifier.issuept. 2en_US
dc.identifier.spage194en_US
dc.identifier.epage203en_US
dc.publisher.placeGermanyen_US
dc.description.otherWorkshop on Computer Vision in Vehicle Technology: From Earth to Mars (CVVT:E2M 2010) in conjunction with the 10th Asian Conference on Computer Vision (ACCV 2010), Queenstown, New Zealand, 8-12 November 2010. In Lecture Notes in Computer Science, 2011, v. 6469 pt. 2, p. 194-203-
dc.identifier.scopusauthoridPan, G=36844530700en_US
dc.identifier.scopusauthoridWong, KYK=24402187900en_US

Export via OAI-PMH Interface in XML Formats


OR


Export to Other Non-XML Formats