an experimental study on blinking detection via eeg signals forhuman-robot interaction purposes based on a spherical 2-dof parallel robot

Blinking and eye movement are one of the most important abilities that most people have, even peoplewith spinal cord problem. by using this ability these people could handle some of their activities such asmoving their wheelchair without the help of others. one of the most important fields in human-robotinteraction is the development of artificial limbs working with brain signals. the purpose of this projectconsists in detecting blinking and left and right eye movements via electroencephalogram signals which isalso useful in determining sleepiness. moving window, fast fourier transform and wavelet transform arethe methods which are used in this project. obtained results reveal that the most reliable method to recognizeblinking and eye movements is fast fourier transform since it is almost insensitive and invariant to thedefined variables. after recognition of blinking and eye movement, these actions are transferred to a robotwhich is called the spherical 2-dofparallel robot, built first at the laval university.