Everyone who wears glasses can relate to the experience of a poor fitting frame. While talking with some friends about their bad experiences with eyewear, we thought of some solutions to make eyewear fitting more accurate and individualized.
What it does
Eye to Eye takes a thermal image of someone's face (and some notes on their preferences). This information is used to process an ideal frame for them which is then searched from an existing collections of sunglasses from retailers (sunglasses hut, etc.). If a match is found, the app returns a price estimate to the customer and the option to ship the eyewear to their address.
How I built it
Backend is made in Java, pipes are dropbox API, front end is HTML5 and CSS.
Challenges I ran into
There wasn't enough Android FlirOnes, so we had to build our app with the input from the iOS FlirOne but the backend on java. This meant some complicated piping from the iPhone to the backend than back to the front-end.
How it works
FlirOne thermal imaging is passed to the backend which breaks it down into 3 sections. We gather the most intense heat points (reds) and take an average to draw contours (specifically in the eyes and nose). These lines are then compared with a table to find recommended fit.
What's next for Eye to Eye
Eye to Eye can use the imaging and data to produce a very personalized eyewear model which customers can then customize (change color, shape, etc). 3D-printing could then print out their customized frames which the customer can walk out of the store with.