The problem Smart Naka solves

The problem we are solving is Assam Police’s - Smart Naka

Currently police have to manually check cars at Naka to determine the car is stolen or not.

We trained a model with OpenCV & YoloV5 to make a accurate Number Plate Detection Model. It will send push notifications to nearby police station Police mobile apps to notify them.

We also added a Missing Car Diary Feature where people can add their missing Car Details.

Also We can get details of any vehicle with its vehicle No. which will help the on-field police officer.

There is also a filter Stolen Car details Feature. By selecting police station and number of days, You will get all the stolen car history occurred in the timespan.

This is the problem we are solving and we are really confident about it.

Challenges we ran into

• Extracting text from the scanned image.

  After building a number plate detection model, the main challenge was to extract images from the detected number plate in forecast conditions. Dealing with this, we used OpenCV and Deep learning powered libraries to process the image in such that number can easily be extracted from it.

• Configuring AWS architecture.

  As the whole process involves a lot of computation, it was really challenging to optimize the architecture so that it processes fast and cost efficient.