Automatic Number Plate Recognition: Team THE CORE

Prerequisite packages

Opencv:-

For windows:

pip install opencv-python

For linux:

sudo apt install python3-opencv

Numpy:-

For windows:

pip install numpy

For linux:

sudo apt install python-numpy

---

Code & Files

Our project performs the following steps:

• Frame extraction from video - Done by videoframe.py.
• License plate extraction - Done by plateDetect.py.
• License plate Recognition - Done by detectchar.py.
• Creates a folder wherein the extracted license plate number is stored.

How to run our project:

• Download the "weights" file folder from https://drive.google.com/open?id=1298i7Ocyq943klu08WKEIlLAimIfzTm3
• Add the weights folder to ANPR
• The test video should be added in the "Input" folder.
• Open the terminal.

cd ANPR

python videoframe.py

• Provide the proper filename with extension when it prompts
• Check output in the results folder

---

In order to enhance the computational perfromance a GPU setup is preferrable as it is 500 times faster than CPU