The problem Traffic flow optimization solves

Traffic congestion is becoming one of the major issues with increasing population and automobiles in the cities. Traffic jams not only cause extra delay and stress for the drivers but also increase fuel consumption and air pollution.

In an attempt to reduce traffic congestion, we developed an improved traffic management system in the form of a Computer Vision-based traffic signal controller that can automatically adapt to the traffic situation at the traffic signal. The system sets the green signal time adaptively according to the traffic density at the signal and ensures that the direction with less traffic is allotted a green signal for a shorter duration of time.

Moreover the system generates all the details of detected vehicles and those details are stored on a database. So that these data collections can be used for making intelligent decisions and better traffic rules. The collected data can also be used for paid distribution of the data to online GPS/map providers.

People can get benefited from this system through many ways, they can get updates about the high traffic routes and can re-route their journry and hence saving their precious time.

Challenges we ran into

While making it we faced many difficulties while building the logic for our system.

We have used YOLOv3 since it is very effective on low resolution images (We have tried to make it as much as feasible).

Major challenge we faced was performing the camera switching in our python file. It was a tedious task to calculate approximate time to switch detecting cameras. And storing the data in database along side performing the detection of vehicles was a bit tricky.

While making the simulation python file (the desktop application) , we faced lot of errors in making it and a lot of logic building was required in making it. It took us a lot of time in making the simulatin work possible.

Then another challenge we faced was to run both the python files together and to get a time variable from one python file to another python file. And we were able to solve it using forming a JSON format file.

We got help from the mentors as well , they were very helpful and guided us wherever we got stuck.

While making the BI Dashboard also it was tricky to display all the data in a summerised manner. But eventually we were able to make it!

We learned a lot while making this project.