License Plate Recognition App

This program recognizes the license plate number from a user provided image (if one exists).

Setup

To set up this project, first clone the repository. Then create a virtual environment and activate it:

python -m venv lprenv
source lprenv/bin/activate  # On Windows use `lprenv\Scripts\activate`

It will install the required packages inside the virtual environment.

Running the program

1. First run the following command to start a production server at port 8080 (default configuration). The server will also listen to all IP addresses on the host by default.

waitress-serve app.main:app

2. Once the server is up and running, you can send a request to the server with an image file:

curl -X POST -F "image=sample-test-car.png" http://127.0.0.1:8080/upload

The above sample request sends the image sample-test-car.png as the payload to the /upload POST API.

The API will return the following response:

whole_image_plate_number: "VIPER"

The Jupyter Notebook file /app/license-plate-recognition.ipynb delineates the dataset, methodology, configuration parameters and experimental results.