A Novel Human Computer Interface based on Electrooculogram Signal for Smart Assistive Robots

Abstract

Assistive robots play an increasingly important role in the lives of people with disabilities. More and more research efforts are being focused on using Electrooculogram (EOG) signals in driving assistive robots. This paper presents a novel human computer interface based on EOG signal for controlling assistive robots. Firstly, this paper introduces the generation mechanism of EOG, collection method and experimental scheme. Then analyzes the principle of autoregressive spectral entropy feature extraction algorithm, procedures and extraction results in detailed. Next applies the support vector machine based on multi-classifier model to identify the EOG signal with three-dimensional feature vectors. Finally, a deterministic finite state machine which uses the support vector machine recognition result as a symbol input sequences is designed to express the transfer rule of different control commands. The experimental result shows that the sequence of the EOG signal made by the user can be correctly identified and output the corresponding command.