Citation

Farmer ME, Jain AK. IEEE Trans. Vehicular Tech. 2007; 56(1): 60-80.

Copyright

(Copyright © 2007, IEEE (Institute of Electrical and Electronics Engineers))

DOI

unavailable

PMID

unavailable

Abstract

In this paper, the authors describe their research into a method of airbag suppression dependent upon an occupant classification and tracking system in order to avoid airbag deployment when it could potentially be harmful as in the case of small children or if an adult is situated too close to the steering wheel. The system is visually-based and uses a grayscale camera real-time that uses commercially available hardware. Classification is accomplished through a background decorrelation system using background subtraction. This system has a 98 percent accuracy reading on single static images, although it performs with higher accuracy for infants than for adults. Occupant tracking is accomplished using a Kalman filter that tracks discontinuities naturally present in the suppression application. Authors note that the system needs to be improved in terms of its classification of adults.