Zur Kurzanzeige

dc.contributor.authorDolezel, Petr
dc.contributor.authorSkrabanek, Pavel
dc.contributor.authorStursa, Dominik
dc.contributor.authorBaruque Zanón, Bruno 
dc.contributor.authorCogollos Adrián, Héctor 
dc.contributor.authorKryda, Pavel
dc.date.accessioned2022-11-02T13:50:03Z
dc.date.available2022-11-02T13:50:03Z
dc.date.issued2022-09
dc.identifier.issn1877-7503
dc.identifier.urihttp://hdl.handle.net/10259/7113
dc.description.abstractImplementations of person detection in tracking and counting systems tend towards processing of orthogonally captured images on edge computing devices. The ellipse-like shape of heads in orthogonally captured images inspired us to predict head centroids to determine positions of persons in images. We predict the centroids using a fully convolutional network (FCN). We combine the FCN with simple image processing operations to ensure fast inference of the detector. We experiment with the size of the FCN output to further decrease the inference time. We compare the proposed centroid-based detector with bounding box-based detectors on head detection task in terms of the inference time and the detection performance. We propose a performance measure which allows quantitative comparison of the two detection approaches. For the training and evaluation of the detectors, we form original datasets of 8000 annotated images, which are characterized by high variability in terms of lighting conditions, background, image quality, and elevation profile of scenes. We propose an approach which allows simultaneous annotation of the images for both bounding box-based and centroid-based detection. The centroid-based detector shows the best detection performance while keeping edge computing standards.es
dc.description.sponsorshipThe work was supported from ERDF/ESF “Cooperation in Applied Research between the University of Pardubice and companies, in the Field of Positioning, Detection and Simulation Technology for Transport Systems (PosiTrans)” (No. CZ.02.1.01/0.0/0.0/17_049/0008394).es
dc.format.mimetypeapplication/pdf
dc.language.isoenges
dc.publisherElsevieres
dc.relation.ispartofJournal of Computational Science. 2022, V. 63, 101760es
dc.rightsAtribución 4.0 Internacional*
dc.rights.urihttp://creativecommons.org/licenses/by/4.0/*
dc.subjectPerson detectiones
dc.subjectFully convolutional networkses
dc.subjectPerformance measurees
dc.subjectEdge computinges
dc.subjectComputer visiones
dc.subject.otherInformáticaes
dc.subject.otherComputer sciencees
dc.subject.otherFotografíaes
dc.subject.otherPhotographyes
dc.titleCentroid based person detection using pixelwise prediction of the positiones
dc.typeinfo:eu-repo/semantics/articlees
dc.rights.accessRightsinfo:eu-repo/semantics/openAccesses
dc.relation.publisherversionhttps://doi.org/10.1016/j.jocs.2022.101760es
dc.identifier.doi10.1016/j.jocs.2022.101760
dc.relation.projectIDinfo:eu-repo/grantAgreement/Ministerstvo školství, mládeže a tělovýchovy České republiky//CZ.02.1.01%2F0.0%2F0.0%2F17_049%2F0008394
dc.journal.titleJournal of Computational Sciencees
dc.volume.number63es
dc.page.initial101760es
dc.type.hasVersioninfo:eu-repo/semantics/publishedVersiones


Dateien zu dieser Ressource

Thumbnail

Das Dokument erscheint in:

Zur Kurzanzeige