This is our web-app that provides location data on stolen/Amber Alert/Silver Alert cars
This is our firebase database that keeps track of important license plate numbers and records the time and location of vehicle sightings.
An image from the live demo we did.
Dash home screen
Dash recording screen - recognizes any text on the live camera feed and displays it.
Every year, first-responders issue a significant amount of alerts to save abducted children, senior citizens with degenerative diseases, and and stolen vehicles. With an increasing amount of Americans using dash cams, we thought it'd make sense to combine text-recognition computer vision with these cameras in order to increase the success rate of the vehicles associated with these alerts being found.
What it does
This project is broken up into two components:
An intelligent dash-cam android app built using Java and Android Studio that uses text-recognition to scan the live camera feed for a set of license plates. If found, it reports its location, the license plate, and a time stamp to our database.
A full-stack web app intended for first-responders that pulls the reported data from a Firebase database and displays pertinent location data for each vehicle being searched for.
How we built it
We built the android app using Android Studio, Java, and an OCR text-recognition library. We used Firebase for our backend and communicated with it from the mobile app and web app using custom-built HTTPS functions. We used Angular and Google Maps API for our front-end to create a slick web app for first-responders.
Challenges we ran into
No one on our team had much experience with Android Studio, so we often spent a significant amount of time debugging trivial issues and running into problems with our dependencies and version control. Also - staying up until 6:30 am drastically reduces critical thinking skills.
Accomplishments that we're proud of
We're extremely proud of the wide range of technologies our project used. We've never combined web development, app development, database functions, and computer vision into a single product and we believe we build a cohesive and impact-driven application using all of these.
What I learned
We learned that spending a significant time planning the project and delegating tasks allows us to accomplish significantly more and create a much more well-fleshed-out product. We also learned that a 1-2 hour power nap can re-energize us and revitalize our brains :)
What's next for Dash
We believe that adding functionality to the app where we can recognize a car crash or emergency and notify first-responders such as ambulances could allow them to get to the scene faster and potentially save more lives.