Gesture Recognition Using an EEG Sensor and an ANN Classifier for Control of a Robotic Manipulator

In recent years, electroencephalography (EEG) has gained popularity in the field of brain-computer interface (BCI). Current applications of BCI include control of prosthetics and robotic systems. In this project, the goal is to acquire and record EEG signals generated by human subjects performing specific facial gestures, and use them to control a robotic hand. Six facial gestures have been selected for this project: smile, raise eyebrows, look right, look left, hard blink, and blink. Once the signals were collected, a classification system based on artificial neural network (ANN) was designed. The classification system was able to recognize and differentiate each gesture with an accuracy of 98% for signals from a single person, and 75% for signals from multiple persons. The EEG signals were acquired using an Emotiv EPOC headset that has 14 sensors. This headset was selected mainly because its portability, affordable cost compared to similar products in the market, and it is easy to place on the subject’s head. The ultimate purpose of this research is to use the classification system output to send control signals to a robotic hand that has been designed and built in our research lab. In this paper, the data collection, data conditioning, design, testing and results of the classification system are provided in detail.