EyeBoard is a head-mount low-cost eye-tracker built in the Lab. It uses two micro-cameras, one aiming at the eye and a panoramic front-oriented one. We use illuminate with infra-red light the target eye and apply the correspondig wave lenth filter to the camera.
Our pupil tracking algorithm combines position accuracy, robustness and low processing load making it suitable for FPGA or DSP embedding. Current research focuses on computer-accesability applications for handicapped people and visual-attention models.
Videos show virtual keyboard writing applications and eye tracking in different scenarios such as reading, video watching and video game playing.
Master Thesis, Universidad de Extremadura, Escuela Politécnica de Cáceres. Enrique Turégano Pedruelo. «EyeBoard: Un Periférico Alternativo Visual».