A SmartWatch that is capable of continously identifiying and authroizing a user based on their heart rythm (not beats per minute but the actual shape of the heart beat -the wave features- since these are unqiue and aren't affected by exercise/mood).
The actual stuff in this repo: (Just check the final report pdf!) The actual work didnt make it past the prototype version. Think of it as a proof of concept at the time of publishing. As is, the work is offloaded to a laptop and the watch is just collecting data. Can be easily incorporated into a product. The Machine Learning algo can run on a phone easily (its linear) or even the cloud or a router or the watch itself if it has the CPU for it.
Three Leads: Can be replaced by only two with a comparator circuit. One will be on the back of the watch so it touches your left hand and the other will be on top of the watch so that when you use your right hand to touch it it takes the reading and verifies its you.
Researched and developed in McMaster University - Electrical & Computer Engineering Department (ECE). 2016-2017