Assistive Smart Home Environment using Head Gestures and EEG Eye Blink Control Schemes

Abstract

Field of Assistive Smart Homes has emerged with the aim of enabling the physically challenged, the elderly or those with constraint motion and to restore their capability of performing necessary daily life tasks by providing required assistance using modern technological tools. Objective of this work is to study possibility of using various available technological tools to enable such people perform independently in main stream life by giving them control of their environment and movement. The said objective is achieved using hybrid physiological gestures, such as, head movement and eye blinks, as even quadriplegic patients can perform these gestures easily. The orientation or movement of head is sensed by a head set embedded with an Inertial Measurement Unit (IMU) and Linear Discriminant Analysis (LDA) is used to recognize the intended command. Eye blinks are detected by sensing Electroencephalography (EEG) signals. After pre-processing, EEG signals are classified on the basis of various signal properties and converted into commands. With the combination of head orientation sensing and eye blink signals, a hierarchy of control commands is generated to control lights, fans, security lock and wheel chair movement. Using the prototype headset, the home environment is simulated and verified in Matlab environment and a GUI is designed for ease of user. The results show the feasibility of the designed system in real time, with average system accuracy of approximately 81.48%, making this design a good and reasonably priced choice for implementation in Assistive Smart Homes, especially in developing countries with low per capita income.