|
KMID : 1144120160060040234
|
|
Biomedical Engineering Letters
2016 Volume.6 No. 4 p.234 ~ p.241
|
|
|
A multi levels data fusion approach for an electric wheelchair control
|
|
Taher Fatma Ben
Amor Nader Ben
Jallouli Mohamed
Hammouda Aicha Ben
Dghim Olfa
|
|
|
Abstract
|
|
|
Purpose: Assistive technology has allowed providing valuable assistance to people with special needs. Several studies have targeted “smart” wheelchairs. However, these prototypes are not yet commercialized, mainly for security reasons. The objective of this study is to ameliorate the control success rate and fulfill maximally the security constraint.
Methods: In this paper, we present a multi-levels data fusion technique. Three subsystems operate separately to find the navigation command. These subsystems are based respectively on EEG signals, head position tracking and eye tracking. Their individual decisions are then fused to deliver the final command to the wheelchair.
Results: The proposed technique achieves was able to reach a success rate equal to 94.4%. This rate represents an improvement of 7% compared to single source control techniques.
Conclusions: The results of this study demonstrate the ability of data fusion to improve the success rate and to avoid critical situations such as total breakdown in case of sensor failure.
|
|
|
KEYWORD
|
|
|
Electroencephalogram (EEG), Head movement, Eye tracking, Data fusion, EPW control, Human-computer interaction
|
|
|
FullTexts / Linksout information
|
|
|
|
|
|
Listed journal information
|
|
|
|