Automatic License Plate Detection in Python

Developed by Anne-Sophie Bollmann, Susanne Klöcker, Pia von Kolken and Christian Peters

We used Tensorflow, OpenCV and Tesseract to build an automatic license plate detection system in Python.

Usage

Open a command prompt, navigate to the src directory and type the following:

python license_plate_detection.py --visualize [path to .jpg image]

You will see something like this:

Installation

This project requires Python 3.8.7.

1. Clone this repository to your local machine:

   git clone https://github.com/cxan96/license_plate_detection.git

2. Navigate into the cloned repository and create a virtual environment:

   cd license_plate_detection/
   python -m venv venv
   ./venv/scripts/activate

3. Install the project requirements:

   pip install -r requirements.txt

Congratulations! 🎉🎉 🎉
You are now ready to predict license plates using the script license_plate_detection.py inside of the src directory.

Notes on Tesseract

This project contains a full Tesseract installation for Windows 10. If you are using Linux or Mac, you have to install Tesseract yourself and update the following line in the file src/character_recognition/ocr_pipeline.py:

pytesseract.pytesseract.tesseract_cmd = "[path to tesseract executable]"