Human posture estimation using voxel data for "smart" airbag systems: issues and framework

Author

Cheng, Shinko Y. ; Trivedi, Mohan M.

Author_Institution

Comput. Vision & Robotics Res., California Univ., San Diego, La Jolla, CA, USA

fYear

2004

fDate

14-17 June 2004

Firstpage

84

Lastpage

89

Abstract

This paper examines the feasibility of a multi-camera voxel based occupant posture estimation system. Several new considerations are made to allow this tested human body modeling system to work reliably in the passenger seat of a vehicle, including camera position, segmentation, and body modeling with voxel reconstructions all from a constrained 4 camera setup. To describe occupant posture, a partial human body model consisting of a head and torso is proposed. The accuracy of the estimation of this model is compared against ground-truth.

Keywords

automobiles; image reconstruction; image segmentation; physiological models; safety devices; video cameras; camera position; feasibility; human body modeling system; human posture estimation; image segmentation; multicamera voxel; occupant posture estimation; passenger seat; smart airbag systems; vehicle; voxel data; voxel reconstructions; Biological system modeling; Calibration; Cameras; Computer vision; Head; Humans; Image reconstruction; Robot vision systems; Torso; Vehicles;

fLanguage

English

Publisher

ieee

Conference_Titel

Intelligent Vehicles Symposium, 2004 IEEE

Print_ISBN

0-7803-8310-9

Type

conf

DOI

10.1109/IVS.2004.1336360

Filename

1336360