Reconnaissance 3D des Gestes pour l’Interaction Naturelle Homme Robot

Abstract : In this paper we propose a 3D dynamic gesture recognition method for human robot interaction (HRI) based on depth information provided by a Kinect. The body is tracked using the skeleton algorithm provided by the Kinect SDK. The main idea of this work is to compute the angles of the upper body articulations which are active when executing gesture. The variation of these angles will be used as inputs for a Hidden Markov Models (HMMs) in order to recognize the dynamic gestures. Results demonstrate that our method is very robust ; it does not need much preprocessing steps and is not influenced by environment conditions like illumination changes and scene complexity.
Document type :
Conference papers
Complete list of metadatas

Cited literature [26 references]  Display  Hide  Download

https://hal.archives-ouvertes.fr/hal-01177441
Contributor : Frédéric Davesne <>
Submitted on : Monday, July 20, 2015 - 3:15:52 PM
Last modification on : Monday, October 28, 2019 - 10:50:21 AM
Long-term archiving on: Wednesday, October 21, 2015 - 10:16:31 AM

File

Hajar_HIYADI_ORASIS.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-01177441, version 1

Citation

Hajar Hiyadi, Fakhr-Eddine Ababsa, El Houssine Bouyakhf, Christophe Montagne, Fakhita Regragui. Reconnaissance 3D des Gestes pour l’Interaction Naturelle Homme Robot. 15ème édition des journées francophones des jeunes chercheurs en vision par ordinateur (ORASIS 2015), Jun 2015, Amiens, France. (elec. proc.). ⟨hal-01177441⟩

Share

Metrics

Record views

291

Files downloads

200