Presentation 1997/6/20
Keyframe Tracking and Prediction of Human Body in 3D Motion from A Monocular Image Sequence
Yoshito Ohta, Takashi Yamagiwa, Masanobu Yamamoto,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) The model-based approaches for tracking of human bodies in image sequences can be categorized into two types; fitting model to body frame by frame, and integration of estimated pose displacements frame by frame after the model fitting at the initial frame. The latter has an inherent drawback as integration of tracking errors while the one has a grate advantage as small computational efforts compared with the former. This paper proposes a technique which can correct tracking errors by feedback from fitting model to body at several key-frames. This paper also proposes another technique which can predict human pose in occlusion by Kalman filter with AR model. We show successful tracking of an athletic motion at the Atlanta Olympic using the both techniques.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Tracking / Key-frame / Kalman filter / AR model / Relaxation
Paper # PRMU97-53
Date of Issue

Conference Information
Committee PRMU
Conference Date 1997/6/20(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To Pattern Recognition and Media Understanding (PRMU)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Keyframe Tracking and Prediction of Human Body in 3D Motion from A Monocular Image Sequence
Sub Title (in English)
Keyword(1) Tracking
Keyword(2) Key-frame
Keyword(3) Kalman filter
Keyword(4) AR model
Keyword(5) Relaxation
1st Author's Name Yoshito Ohta
1st Author's Affiliation Department of Information Engineering, Niigata University()
2nd Author's Name Takashi Yamagiwa
2nd Author's Affiliation Department of Information Engineering, Niigata University
3rd Author's Name Masanobu Yamamoto
3rd Author's Affiliation Department of Information Engineering, Niigata University
Date 1997/6/20
Paper # PRMU97-53
Volume (vol) vol.97
Number (no) 113
Page pp.pp.-
#Pages 8
Date of Issue