We were inspired to build something that could help people whose interests are not being represented by the engineering community. According to a recent study, more than 1% of American people suffer from some form of paralysis, more than 50,000 of whom have a condition known as Locked-in Syndrome. These people are awake, able to think and observe everything around them, and yet they can do nothing to interact with their environment beyond slight movements of the eyes. They need a voice, and we felt obligated to give them one.
What it does
OSIRIS is an accessibility interface designed to allow people with minimal control over their bodies to be able to navigate and use computers. By utilizing computer vision technologies, OSIRIS tracks the user's eyes allowing them to control the mouse position on their computer and allowing them to click and execute different mouse functions by executing blink-based commands.
How we built it
We built OSIRIS using Python and OpenCV. The application is designed for macOS, but could be easily transferred to Windows and Linux, potentially even mobile.
Challenges we ran into
Some of the biggest challenges that we ran into were with utilizing the OpenCV libraries with C++ (Which was the language OSIRIS was intended to be built with). As result of these difficulties, we migrated OSIRIS to Python. Another issue that we came across was working to make the necessary calculations to accurately track the user's eyes in dynamic conditions, as well as translating those movements into mouse behaviors.
Accomplishments that we're proud of
We are extremely proud of how effective the system is given the time we had to build it, and some of the unavoidable circumstances that brought on challenges to our efforts.
What we learned
We learned a lot about several extremely intriguing concepts behind Computer Vision technology that allow for the mathematical analysis of 2-Dimensional image data and being able to then extrapolate incredulous amounts of useful information from it.
What's next for OSIRIS
We plan to work to make OSIRIS more effective in varying light conditions and variables that reduce with it's effectivity. We want to make this system accessible to anyone who may need it, with nothing more than access to a computer and a webcam.