Abstract:
Detecting human eyes movement is very important in different aspects. One of them is its usage for communication, especially for people with amyotrophic lateral sclerosis (ALS) and other locked-in and paralysis diseases. ALS patients in late stages lose control of their muscles but they almost maintain control of their eye muscles. Therefore, the only remaining way for communication is their eyes. Nowadays, several commercial and non-commercial projects develop a way to make eye movement tracking more applicable, but they usually offer very expensive solutions. Currently, most of eye tracking systems work based on infrared cameras, which are not commonly available and may even be harmful for eyes. For Persian patients, there is not any special solution at all. The proposed system uses just a normal webcam. For more simplicity, calibration-free algorithm is utilized to identify eye and its movement. A special user interface helps patients to type whatever they want. It is integrated with a pervasive algorithm that scores the most exploited words during the time and stores them in a database for using them as suggestions.
Materials:
| Presentation: PDF (English/Persian) | Source code/Executable file/Additional files |