This project proposes to build human-computer interface based on eye movement detection.
Abstract
This project proposes to build human-computer interface based on eye movement detection. As the tested subject eyes move on a horizontal scale the marker in the implemented software will follow his observation point, showing the current point on which the subject gazes. The proposed system includes two parts. The first is the PC-EOG 1200 device which obtaines the signals generated by the eyes (EOG). The second part is the software, implemented in the project – Horizontal Eye Tracker (HET) which makes the detection and the visual presentation of the current observation point.
The EOG Aplications
One of the most developing areas in Engineering which utilizes the extensive research in medicine is the Biomedical Engineering. This area seeks to improve our everyday life by applying engineering and medical knowledge together with the growing power of computers. Developing a convenient Human-Machine interface has been a serious engineering challange in the last few decades. This project explores the possibility of control machines by moving the eyes. The results of this project can be applied in order to help paralyzed people by allowing them to use a PC. This project can also be very useful for advertising, where it’s crucial to understand how people see and detect commercials. Other areas that will gain from Human-Machine interface are medicine, interactive computer games, testing of subject’s responses and attention in simulators for training military and law enforcers.
Improving advertisement with eye-movement studies
Electro-Oculography (EOG)
The electrical potentials around the eyes arise due to the permanent potential difference of between 10 and 30mV that exists between the cornea and the ocular fundus. This is referred to as the cornea-retinal potential with the cornea being positive.
Placing the electrodes for Measuring the cornea-retinal
horizontal and vertical detection potential
The Implementation
The goal of this project was to build an inexpensive eye-movement detecting system. The system gets input from the human tested subject and acts according to it. The input is the electronic signals produced by moving eyes (EOG). The recording of the EOG was done using the PC-ECG device, which distorts the signal. Thus a Software Inverse System was implemented to cancel the distortion. The final mission of our software is to detect and distinguish between horizontal and vertical eye movements. In addition, the software has to determine the actual place on the computer screen on which the tested subject gazes.
Building the Software Inverse System
Results
The final system that we created was able to detect quite accurately on which of 3 the points (located on a horizontal line) the subject gazes. As a result looking on a specific point on a screen caused it to become red. In addition an attempt was made to detect horizontal and vertical moves simultaneously. A satisfactory results were achieved distinguishing between 9 points on the screen.
Block diagram
Horizontal eye-movement detection
Acknowledgment
We are grateful to our project supervisor Eldad Klaiman for his help and guidance throughout this work, and to VISL chief engineer Johanan Erez for his support.