Context-Based Face Control of a Robotic Wheelchair

In this article a method to perform semi- autonomous navigation on a wheelchair is presented, contextual information from the environment as user's habits and points of interest are employed to infer the user's desired destination in a global map. Illogical steering signals coming from the user- machine interface input are filtered out to improve the overall performance of the system. Examples using a face tracking system and voice recognition are presented. The estimation of the user's desired goal is performed employing a Bayesian network. An autonomous navigation system is used to control the wheelchair's low level navigation while the user is just concerned at pointing to the desired destination.