People with severe disabilities, especially spinal cord injury (SCI), encounter many difficulties in communicating with computer systems, because their problem is generally not addressed in computer programs.
With the advent of technology and the move toward Internet of Things (IoT), a new technique has been developed by combining an eye tracking system with IoT to create a platform for the use of SCI patients.
Eye tracking methods include invasive, non-invasive and semi-invasive techniques. A non-invasive method in which the eye movement is traceable in visible or infrared light is used in this research. An algorithm is designed and developed that allows an IoT-based connection providing higher speed and precision compared with previous algorithms. Therefore, this technique is novel in application of eye tracking for issuing control commands in an IoT-based smart home by individuals with SCI or other physical disabilities. The total command receipt, confirmation, and execution time in this system is less than 10 seconds. The optical conditions in this research were 5 ~ 300 lux and 2 ~ 6 lux in invisible and infrared light processing, respectively. It produced 97% correct responses with the help of statistical weighted averaging and elimination of bad data. This system uses the Arduino Mega 2560 board in IoT part, and MATLAB software in feature-based face detection processing.
This methodology improved the quality and accuracy of gaze tracking due to the following reasons: enhanced image clarity, use of two-eye processing, and use of statistical weighted averaging, as well as the use of both visible/infrared lights to reduce the sensitivity to the ambient light. This system has the advantage of having no sensitivity to age, gender, hairstyle, beard, mustache, and veil, and the possibility of producing output for several devices.
The proposed system is reliable for designing a smart home for SCI patients.