We always found other people dangerous after got hurt by them. Although this causality seems reasonable, it never works in favor of innocent people like us. In this project, we are trying to reverse this causality, break this status quo and make people feel safer.
What it does
Our system could detect people who steps in front of your door, record their facial data, save them in database, and then allow user than determine who are friends and who are suspicious.
How I built it
In our server and web side, we use websocket, Flask, D3.js, Mongo DB. For hardware, we use raspberry pi combined with PiCamera, ATmega controllor, motion sensor, ultrasonic sensors and relay.
Challenges I ran into
The raspberry pi cannot handle sensor's data fast enough, then caused huge delay in the system. We solved this problem by adding another MCU in the system to help faster signal condintioning process.
Accomplishments that I'm proud of
We are really happy that our system could work!
What I learned
We learned how to work with websocket and how to do facial recognization.