PeekAWho

• 

• 

  Notification

• 

  Facial Recognition

Inspiration

Sometimes people stop by your dorm room, apartment, or house looking for you but you aren't there but you don't know that they were there. There should be a way of fixing that.

What it does

Our project is in its very early stages. It currently is a program that notifies a user when their friends stop by. It tells the user who exactly stopped by using facial recognition and when they stopped by. It currently uses a webcam now but in the future it will use a camera attached to a raspberry pi. For demo purposes, the notifications currently utilize the standard Microsoft Windows notifications.

Operation

To use, the server is run on the user's computer. The client is then run on the device attached to the camera. The client upon initial startup will read a folder of faces to "learn" them. When one of those faces appears at the camera, a notification is sent to the user stating that person's name, and that they are at the front door. Unknown faces will prompt no response. A known face has to be absent from the camera's view for at least 5 minutes to prompt another notification to be sent to the user, to prevent notification spamming.

How we built it

The program is written in python. We used the facial recognition and OpenCV python libraries as well as expanded our knowledge of the standard python libraries. We use sockets to send information to the users server.

Challenges we ran into

We ran into a problem that became our biggest roadblock. Figuring out the installation and implementation of OpenCv and Face Recognition. We encountered countless improperly specified paths, build errors relating to different Python versions, and even broken virtual machines. This issue took up the majority of the time we spent on the project, and quick and efficient progress was made with the remaining time.

Accomplishments that we're proud of

Actually being able to run the Face Recognition demo for the first time. When the webcam saw our faces and a notification on a different computer informed us of that we were present at the door.

What we learned

One team member learned about sockets. Another learned about the difficulty of compiling your own programs and setting paths. We also are beginning to learn more and more about Face Recognition and OpenCV.

What's next for PeekAWho

Moving the program to a Raspberry Pi with a camera attached. This way it can be attached to the peep hole on a door. Make a native application for mobile devices that will notify the user and/or possibly integrating into another preexisting notification system. Evolving from sockets to a different form of data transfer.

Built With

• python