Cargando…

GrabCut-Based Human Segmentation in Video Sequences

In this paper, we present a fully-automatic Spatio-Temporal GrabCut human segmentation methodology that combines tracking and segmentation. GrabCut initialization is performed by a HOG-based subject detection, face detection, and skin color model. Spatial information is included by Mean Shift cluste...

Descripción completa

Detalles Bibliográficos
Autores principales: Hernández-Vela, Antonio, Reyes, Miguel, Ponce, Víctor, Escalera, Sergio
Formato: Online Artículo Texto
Lenguaje:English
Publicado: Molecular Diversity Preservation International (MDPI) 2012
Materias:
Acceso en línea:https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3522968/
https://www.ncbi.nlm.nih.gov/pubmed/23202215
http://dx.doi.org/10.3390/s121115376
_version_ 1782253147605958656
author Hernández-Vela, Antonio
Reyes, Miguel
Ponce, Víctor
Escalera, Sergio
author_facet Hernández-Vela, Antonio
Reyes, Miguel
Ponce, Víctor
Escalera, Sergio
author_sort Hernández-Vela, Antonio
collection PubMed
description In this paper, we present a fully-automatic Spatio-Temporal GrabCut human segmentation methodology that combines tracking and segmentation. GrabCut initialization is performed by a HOG-based subject detection, face detection, and skin color model. Spatial information is included by Mean Shift clustering whereas temporal coherence is considered by the historical of Gaussian Mixture Models. Moreover, full face and pose recovery is obtained by combining human segmentation with Active Appearance Models and Conditional Random Fields. Results over public datasets and in a new Human Limb dataset show a robust segmentation and recovery of both face and pose using the presented methodology.
format Online
Article
Text
id pubmed-3522968
institution National Center for Biotechnology Information
language English
publishDate 2012
publisher Molecular Diversity Preservation International (MDPI)
record_format MEDLINE/PubMed
spelling pubmed-35229682013-01-09 GrabCut-Based Human Segmentation in Video Sequences Hernández-Vela, Antonio Reyes, Miguel Ponce, Víctor Escalera, Sergio Sensors (Basel) Article In this paper, we present a fully-automatic Spatio-Temporal GrabCut human segmentation methodology that combines tracking and segmentation. GrabCut initialization is performed by a HOG-based subject detection, face detection, and skin color model. Spatial information is included by Mean Shift clustering whereas temporal coherence is considered by the historical of Gaussian Mixture Models. Moreover, full face and pose recovery is obtained by combining human segmentation with Active Appearance Models and Conditional Random Fields. Results over public datasets and in a new Human Limb dataset show a robust segmentation and recovery of both face and pose using the presented methodology. Molecular Diversity Preservation International (MDPI) 2012-11-09 /pmc/articles/PMC3522968/ /pubmed/23202215 http://dx.doi.org/10.3390/s121115376 Text en © 2012 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution license (http://creativecommons.org/licenses/by/3.0/).
spellingShingle Article
Hernández-Vela, Antonio
Reyes, Miguel
Ponce, Víctor
Escalera, Sergio
GrabCut-Based Human Segmentation in Video Sequences
title GrabCut-Based Human Segmentation in Video Sequences
title_full GrabCut-Based Human Segmentation in Video Sequences
title_fullStr GrabCut-Based Human Segmentation in Video Sequences
title_full_unstemmed GrabCut-Based Human Segmentation in Video Sequences
title_short GrabCut-Based Human Segmentation in Video Sequences
title_sort grabcut-based human segmentation in video sequences
topic Article
url https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3522968/
https://www.ncbi.nlm.nih.gov/pubmed/23202215
http://dx.doi.org/10.3390/s121115376
work_keys_str_mv AT hernandezvelaantonio grabcutbasedhumansegmentationinvideosequences
AT reyesmiguel grabcutbasedhumansegmentationinvideosequences
AT poncevictor grabcutbasedhumansegmentationinvideosequences
AT escalerasergio grabcutbasedhumansegmentationinvideosequences