The iPhone X gives you the ability to unlock your phone with your phone. We give you the ability to open a door with your face and enter your house. If you tend to lost your keys, you don't have to worry anymore, because now you can open the door with your face.
What it does
The door will open if it recognizes you as the owner of the house, it will snap close if you are an intruder and not the owner of the house.
How we built it
The door is built using a cardboard and controlled by using a servo motor with Raspberri Pi 3. The face recognition is done using OpenCV 3 and LBPH face recognizer. The owner first go on a website built using Flask to enter their name, and snap picture of their face using the webcam to train the data. Then they can stand in front of the door. The picamera will snap a picture of them, and then send to Firebase. If the picture is similar to data, then the door will open and said "Welcome home". If the picture is not of the owner, the door will alert of Intruder and send a text message to the owner's phone to alert them using Twilio API.
Challenges we ran into
We ran into A LOT OF challenges. This is the first time that we work with a Rasperry Pi so we have problem connecting it to Dragonfly. We have to resolve to using a hotspot on our own laptop and use VNC viewer to controlled it from our laptop. The face recognition was also very challenging. We also have some slight problem integrating Twilio API and basically just communicate to the Pi and move the door with our servo motor. Also the cardboard is not very stable so if the door move to fast it will fell off.
Accomplishments that we're proud of
We finally was able to move the door using our computer. We can also recognizer the owner's face and had the computer saying "Welcome home, owner" if it's the owner. We were able to configure the Pi and also worked with OpenCV. We also was able to connect everything together and create a GUI for better user interface using a Flask App.
What we learned
We learn how to run a Raspberri Pi using NOOBS and VNC viewer to run the Pi Camera. We also learn how to open the webcam from a Flask App, and communicate to the Pi using Pyrebase - a Python library wrapper for Firebase. We were able to successfully integrated Twilio API and Windows Speech Engine to send an alert message to the owner of the house if someone is trying to enter their house.
What's next for Biometric Door
Optimize the face recognition and make a better door instead of one out of cardboard.