• We wanted to build something to help combat school shootings and other crimes that take place due to unrecognized people in schools. We decided the best way to go about this was by making a security camera that could alert a user if it detects an unrecognized individual.

What it does

  • School Security is a camera that is trained to recognize 3 cases: verified students, unrecognized individuals, and no people. It then either rings a bell, sets off an alarm, or doesn't alert for each case respectively. ## How I built it
  • School Security is built on a Raspberry Pi running a Teachable Machine. As discussed in the video, it is connected to a webcam (the camera), a speaker, as well as a status light. Using python and javascript, the teachable machine is able to constantly differentiate between each of the three cases, performing the specified action (ringing a bell, setting off an alarm, or nothing) for each case. ## Challenges I ran into
  • Since both of us are beginners with ML, we had to learn most of what we used in this project from scratch. We also had to learn how to write Javascript as this was the first web-based project either of us has done. ## Accomplishments that I'm proud of
  • Using ML for the first time, integrating the Teachable Machine on the Raspberry Pi, and creating a standalone product that has no need for an external monitor or computer. ## What I learned
  • How to use ML and Javascript ## What's next for School Security
  • Integrate different systems of notifying a principal/security guard (sms, email, iOS push notification)
  • Test out School Security in schools to see what other features would be useful that could be implemented.

Built With

Share this project: