Designing a low-cost eyeball tracking keyboard for paralyzed people

This paper aims at designing a scanning keyboard that serves people with paralysis. This keyboard works on detection and tracking of the user eyeball movements. The main objective of this research work is to implement a simple and less expensive technique that produces the best possible results with a higher level of accuracy. The design of the system is done in phases; these are eye detection and tracking, followed by classification. The tracking phase incorporates skin color segmentation, black pupil detection and support vector machines. We also explored the role of eye red-channel, eye width and height for eye tracking. A scanning keyboard is developed and tested to work with three eyeball movements with 90% double detection accuracy. The performance of the algorithm is experimentally analyzed and the benefits of the proposed approach are highlighted. Also, the simulation results are presented.