Traffic Watcher

• 

  Map-view that the application opens into. The green dots represent locations of speed measurement.

• 

  The page brought up when a green dot is clicked, which displays the camera number and current average speed in kilometers per hour.

Inspiration...

We knew that traffic was always going to be a problem for lots of people commuting every day. It was also hard to quantify how severe the traffic was on a certain stretch of road. Google Maps could only provide qualitative measurements and so users could only sometimes guess what they would be in for.

What we did...

We decided to build a real-time traffic monitoring system using opencv to analyze car movement on the roads. That way we could get traffic data to users rapidly and accurately. By utilizing Georiga's surveillance camera system, we created a mock of this system and using trained tensorflow models, we could track car movement and ultimately derive the speed of traffic flow.

What we learned...

Going into this project, we knew next to nothing about how to use opencv, or tensorflow, or Firebase and had minimal experience with the tools we used in the final product. We were pushed to learn quickly and adapt existing ideas to our own. By the end, our freshman knowledge had considerably been increased. Everything was a challenge, but an interesting one.

Built With

• android-studio
• firebase
• opencv
• python
• tensorflow