Everyday we commute to and from college. We experience several red lights on the way and some of them are congested due to poor implementation of their management. This problem statement led us to a optimized red light system leading to decongestion of city roads.
What it does
The system analysis a live feed coming from a camera connected to a traffic signal, and a machine learning algorithm is implemented locally for the junction. Number of vehicles are calculated through image preprocessing which optimisis the number of seconds for the green light through the regression model obtained through machine learning. The end user is also provided with an app that works along side his navigation which tells him whether the next light is red or green so he could adjust the speed of his car.
How we built it
A script was created in python importing the library openCV which provides live streaming data. The script pre-processed the stream recognizing the number of vehicles on the road. Website was created which provides the user with the data regarding the traffic light For optimization of the traffic light, machine learning is implemented for a linear regression model
Challenges we ran into
Accomplishments that we're proud of
Usage of openCV which provides live stream and pre-processing the video
What we learned
We learned how to import different libraries in python. Combining python back end with front end .
What's next for Traffic 2.0
The hardware implementation of the software which would optimize the traffic signals in real time. Obtaining the live stream and training the system locally to optimize the duration of the green signal.