Emotion detection for wheelchair navigation enhancement

The goal of this study is to investigate the use of emotion as a braking system for wheelchair navigation. In the first part emotion is detected based on ElectroEncephalography (EEG) technology and emotion induction experiments. Using different techniques for features extraction (Welch and Wavelets), selection (Principal Component Analysis (PCA) and Genetic Algorithm (GA)) and classification (Support Vector Machine (SVM), Multi Layer Perceptron (MLP) and Linear Discriminate Analysis (LDA)), the best combination was assigned to Wavelets-GA-MLP. In the second part, in order to validate the impact of emotion as velocity modulator, a comparison between emotion-based and non emotion-based wheelchair navigation scenarios in a simulated environment was conducted. The assessment was based on four parameters: obstacles hit, navigation path, execution time and outbound points of gaze (POG). While the first two emotion introductions showed better results, this was not the case for the third. These findings can be utilized in order to prescribe a suitable wheelchair according to the subject pathology.