The use of illegal vehicles, such as PMDs, is not properly monitored which is dangerous for pedestrians.

What it does

Detect a variety of vehicles based on footage from CCTVs across Singapore.

How we built it

The core algorithm is Haar Cascade Classifier which is taken from the OpenCV library. The UI was created using Tkinter and web browser layouts were built by html5, js, google maps API, pandas, and matplotlib.

Challenges we ran into

The lack of footage makes accurately detecting vehicles difficult. Communication between Python and javascript files for data visualization.

Accomplishments that we're proud of

Made the detection accurate to a certain extent and managed to build a complete UI. Applied Data Science principles for the data visualization tools like graph and heat map.

What we learned

Basic Computer Vision algorithm to detect objects using Haar Cascade Classifier. Integrate a Google Maps API into our application.

What's next for Vehicle Radar and Data Visualisation

Make the detection more robust by obtaining more data for the Haar Cascade Classifier.

Built With

Share this project: