Apple's recent release of iphone x has demonstrated to the world the practical applications of facial recognition. While the technology has existed for a while, no one had found a way to incorporate it, on a large scale, into the everyday lives of normal people. Coming into this hackathon, our group decided we wanted to explore the applications of facial recognition technology.
What it does
A desktop application that recognizes your face, compares it to a set of authorized user faces, and if it registers a match, unlocks a door.
How we built it
wxPython was used to build a graphical user interface that allows an individual to take their picture and store it into a remote database. The same application then uses Microsoft's Azure Face API to compare a photo of an individual's face to any of the other photo's saved on the data base. If a match is registered, the program executes a Bash script that remotely connects to the dragonboard and executes another python script on the board which activates the servo hardware.
Challenges we ran into
- Integrating GUI
- Connecting to the hardware remotely
- Finding a common programming language
- Executing bash in python
- Using sensor mezzanine
- Uploading code to Arduino Mega
- Mapping ports to the sensor mezzanine
- Setting up MS Azure
- Setting up a static IP address on public network
- Connecting to Wifi
Accomplishments that we're proud of
It works! Managed to get computers to communicate with the board, built a simple GUI for adding photos.
What we learned
How to write python code, RESTful commands, Microsoft Azure, SSH protocols, dragonboards
What's next for Easy In Facial Recognition
If possible it would be great to further develop the program to integrate a mobile application which would allow someone to give people access into their house using the camera. We would also like to add different privileges to different people and make more security considerations.