Watch Dog Logo
Face Recognition Proof
As engineers, we would stay late on campus frequently to finish projects and homework before deadlines. However, with staying on campus so late, we would feel unsafe going home sometimes, even with the blue button system. Keeping this in mind, we were thinking of creating another system that could be implemented alongside the other security systems in situations where someone doesn’t have enough time to call 911 and notify the operator of his or her location and/or he or she cannot reach the blue button or a safer place. That’s when we thought of Watch Dog.
What it does
Watch Dog would add an additional layer of security for college students on campus, especially late at night. It is an app where a person can click a button and send a SMS to WatchDog to call for help and a quadcopter in the sector will come to him or her, record the incident, and send 911 his or her location. Additionally, we have a voice recognition component in case a person cannot click the button quickly enough. In this case, the user could say, “S.O.S” on their phone, and the quadcopter would come to the user.
How we built it
For our project, we initially split up and worked on the different features. Two people were coding in Python on the raspberry pi to have the light turn on, the audio play, and the video record when ran. The other two were working on Android studio to make the app and worked on communication between the raspberry pi and the app. Then we all worked on the communication between the app and quadcopter. We implemented Twilio to activate our app, the Microsoft API to do facial recognition, a servo connection to turn on an activator to represent the motor of the quadcopter, implemented voice recognition through the phone, and debugged the project.
Challenges we ran into
We have difficulty sending the signal between the app and the raspberry pi. We had difficulty using the Microsoft API because the API needed images while we only had video input. Also we were challenged by time to finish our project so the the MavLink connection between the computer and the Pixhop flight connection. Also lack of materials like soldering and tape impaired our work.
Accomplishments that we're proud of
We’re proud of achieving the start of the security system we envisioned. Also, we're happy about getting the communication to work and making the watch dog logo.
What we learned
This project taught us how to better use python as most members have not really used it before.
What's next for Watch Dog
We are going to work on implementing the quadcopter with the code, flight testing, and improving communication to and from the quadcopter and raspberry pi in the future.