Development of an electrooculogram-based human-computer interface for hands-free control of assistive devices

Abstract

The current study proposes the development of an electrooculogram (EOG)-based human-computer interface (HCI) for hands-free control of assistive devices. A commercially available robotic arm was customized and used as a representative assistive device. The EOG signal was acquired in a laptop using the developed EOG data acquisition module (EOG-DAQ). The acquired EOG signals were classified using a novel dynamic threshold algorithm. The control signals were generated by simultaneous events of hall-effect (HE) sensor activation and eye movement detection. This control mechanism was employed to avoid false activation of the assistive device. The transmission of the control signals to the robotic arm was performed using Xbee communication protocol. The performance of the developed system was evaluated by a customized pick-and-place experiment by 10 human volunteers. All the volunteers were able to perform the tasks successfully. The execution time could be reduced with a short training to the volunteers.