Show simple item record

dc.contributor.authorİkizler, N.en_US
dc.contributor.authorDuygulu, P.en_US
dc.date.accessioned2016-02-08T10:02:41Z
dc.date.available2016-02-08T10:02:41Z
dc.date.issued2009-09-02en_US
dc.identifier.issn0262-8856
dc.identifier.urihttp://hdl.handle.net/11693/22633
dc.description.abstractMost of the approaches to human action recognition tend to form complex models which require lots of parameter estimation and computation time. In this study, we show that, human actions can be simply represented by pose without dealing with the complex representation of dynamics. Based on this idea, we propose a novel pose descriptor which we name as Histogram-of-Oriented-Rectangles (HOR) for representing and recognizing human actions in videos. We represent each human pose in an action sequence by oriented rectangular patches extracted over the human silhouette. We then form spatial oriented histograms to represent the distribution of these rectangular patches. We make use of several matching strategies to carry the information from the spatial domain described by the HOR descriptor to temporal domain. These are (i) nearest neighbor classification, which recognizes the actions by matching the descriptors of each frame, (ii) global histogramming, which extends the idea of Motion Energy Image proposed by Bobick and Davis to rectangular patches, (iii) a classifier-based approach using Support Vector Machines, and (iv) adaptation of Dynamic Time Warping on the temporal representation of the HOR descriptor. For the cases when pose descriptor is not sufficiently strong alone, such as to differentiate actions "jogging" and "running", we also incorporate a simple velocity descriptor as a prior to the pose based classification step. We test our system with different configurations and experiment on two commonly used action datasets: the Weizmann dataset and the KTH dataset. Results show that our method is superior to other methods on Weizmann dataset with a perfect accuracy rate of 100%, and is comparable to the other methods on KTH dataset with a very high success rate close to 90%. These results prove that with a simple and compact representation, we can achieve robust recognition of human actions, compared to complex representations. © 2009 Elsevier B.V. All rights reserved.en_US
dc.language.isoEnglishen_US
dc.source.titleImage and Vision Computingen_US
dc.relation.isversionofhttp://dx.doi.org/10.1016/j.imavis.2009.02.002en_US
dc.subjectAction recognitionen_US
dc.subjectHuman motion understandingen_US
dc.subjectPose descriptoren_US
dc.subjectAccuracy rateen_US
dc.subjectAction sequencesen_US
dc.subjectCompact representationen_US
dc.subjectComplex modelen_US
dc.subjectComputation timeen_US
dc.subjectData setsen_US
dc.subjectDescriptoren_US
dc.subjectDynamic time warpingen_US
dc.subjectHistogrammingen_US
dc.subjectHuman poseen_US
dc.subjectHuman silhouetteen_US
dc.subjectHuman-action recognitionen_US
dc.subjectMotion energyen_US
dc.subjectNearest neighbor classificationen_US
dc.subjectRectangular patchen_US
dc.subjectRobust recognitionen_US
dc.subjectSpatial domainsen_US
dc.subjectTemporal domainen_US
dc.subjectTemporal representationsen_US
dc.subjectGesture recognitionen_US
dc.subjectParameter estimationen_US
dc.subjectStatistical testsen_US
dc.subjectHuman form modelsen_US
dc.titleHistogram of oriented rectangles: a new pose descriptor for human action recognitionen_US
dc.typeArticleen_US
dc.departmentDepartment of Computer Engineeringen_US
dc.citation.spage1515en_US
dc.citation.epage1526en_US
dc.citation.volumeNumber27en_US
dc.citation.issueNumber10en_US
dc.identifier.doi10.1016/j.imavis.2009.02.002en_US
dc.publisherElsevier BVen_US


Files in this item

Thumbnail

This item appears in the following Collection(s)

Show simple item record