Electrooculogram based human-machine interface application

In this study, electrooculogram (EOG) based human-machine interface (HMI) application is proposed for partial or completely paralyzed and physically limited people as Amyotrophic Lateral Sclerosis (ALS) patients. In the designed system, EOG signals consisted of vertical and horizontal eye movements were detected by using 5 Ag-AgCl electrodes which were placed around the eye. Interpretable EOG signs were placed acquired through various reinforcement and filtration processes on the signals. Control signals to be used in HMI applications were acquired as a result of digital signal processing of the analog EOG data by the microcontroller unit. After preprocessing step, feature extraction was conducted for left, right, upwards, and downwards eye movements. These values have a direct impact on the performance of the classification step. Artificial Neural Network (ANN), k-Nearest Neighbour (k-NN), and Support Vector Machines (SVM) were used for classification. Acording to the obtained results, ANN, k-NN, and SVM methods performed the classification process with an accuracy rate of 83%, 73%, and 74%, respectively. Simulation results show that the individuals with reduced mobility might successfully communicate with others via their eye movements through this EOG-based HMI system. The measured signals are applied to various amplification and filtering circuits and by this way usable EOG signals are achieved.