Acquisition 3D des gestes par vision monoscopique en temps réel

Abstract : Our topic is 3D human motion capture by real-time monocular vision without marker. We extend a previous work on colour-region based registration of a 3D articulated model of the human body on video sequences (Marques Soares et al., 2004). We describe a video frame-rate implementation for registering image edges with the 3D model occluding edges. The later are extracted using a consumer graphics processing unit (GPU).We give some example results and discuss the precision improvement achieved
Document type :
Conference papers
Complete list of metadatas

https://hal.archives-ouvertes.fr/hal-01378702
Contributor : Médiathèque Télécom Sudparis & Institut Mines-Télécom Business School <>
Submitted on : Monday, October 10, 2016 - 4:21:11 PM
Last modification on : Thursday, September 5, 2019 - 7:02:02 PM

Identifiers

  • HAL Id : hal-01378702, version 1

Citation

David Antonio Gomez Jauregui, Patrick Horain, Fawaz Baroud. Acquisition 3D des gestes par vision monoscopique en temps réel. MajecSTIC 2008 : Manifestation des Jeunes Chercheurs en Sciences et Technologies de l'Information et de la Communication, Oct 2008, Marseille, France. ⟨hal-01378702⟩

Share

Metrics

Record views

35