Driver verification based on handgrip recognition on steering wheel
conference contribution
posted on 2011-01-01, 00:00authored byRonghua Chen, Fenghua She, Jin Wang, Xiangping Sun, Lingxue KongLingxue Kong
This paper presents a novel driver verification algorithm based on the recognition of handgrip patterns on steering wheel. A pressure sensitive mat mounted on a steering wheel is employed to collect a series of pressure images exerted by the hands of the drivers who intend to start the vehicle. Then, feature extraction from those images is carried out through two major steps: Quad-Tree-based multi-resolution decomposition on the images and Principle Component Analysis (PCA)-based dimension reduction, followed by implementing a likelihood-ratio classifier to distinguish drivers into known or unknown ones. The experimental results obtained in this study show that the mean acceptance rates of 78.15% and 78.22% for the trained subjects and the mean rejection rates of 93.92% and 90.93% to the un-trained ones are achieved in two trials, respectively. It can be concluded that the driver verification approach based on the handgrip recognition on steering wheel is promising and will be further explored in the near future.
History
Event
IEEE International Conference of Systems, Man, and Cybernetics (2011 : Anchorage, Alaska)
Pagination
1645 - 1651
Publisher
IEEE
Location
Anchorage, Alaska
Place of publication
Piscataway, N. J.
Start date
2011-10-09
End date
2011-10-12
ISSN
1062-922X
ISBN-13
9781457706523
ISBN-10
1457706520
Language
eng
Publication classification
E1 Full written paper - refereed
Copyright notice
2011, IEEE
Title of proceedings
SMC 2011 : Conference proceeding of the 2011 International Conference on Systems, Man, and Cybernetics