Easy local ALPR (Automatic License Plate Recognition)

This script is a REST API server that uses ultimateALPR-SDK to process images and return the license plate information. The server is created using Flask and the ultimateALPR SDK is used to process the images.

This script is intended to be used as a faster local alternative to the large and resource heavy CodeProject AI software.

Important

The ultimateALPR SDK is a lightweight and much faster alternative (on CPU and GPU) to the CodeProject AI software but it has a few limitations with it's free version:

• The last character of the license plate is masked with an asterisk
• The SDK supposedly has a limit of requests per program execution (never encountered yet) but I have implemented a workaround for this by restarting the SDK after 3000 requests just in case.

Usage

The server listens on port 5000 and has one endpoint: /v1/image/alpr. The endpoint accepts POST requests with an image file in the 'upload' field. The image is processed using the ultimateALPR SDK and the license plate information is returned in JSON format. The reponse follows the CodeProject AI ALPR API format. So it can be used as a drop-in replacement for the CodeProject AI ALPR API.

POST: http://localhost:5000/v1/vision/alpr

Parameters

• upload: (File) The image file to process. (see Pillow.Image.open() for supported formats)

Response

{
  "success": (Boolean) // True if successful.
  "message": (String) // A summary of the inference operation.
  "error": (String) // (Optional) An description of the error if success was false.
  "predictions": (Object[]) // An array of objects with the x_max, x_min, max, y_min bounds of the plate, label, the plate chars and confidence.
  "processMs": (Integer) // The time (ms) to process the image (includes inference and image manipulation operations).
}

Included models in built executable

When using the built executable, only the latin charset models are bundled by default. If you want to use a different charset, you need to set the charset in the JSON_CONFIG variable and rebuild the executable with the according models found here To build the executable, you can use the build_alpr_api.sh script, which will create an executable named alpr_api in the dist folder.

Setup development environment

Install ultimateALPR SDK

Use already built wheel

I have already built the ultimateALPR SDK for x86_64 and ARM64 and included the python3.10 wheel in the wheel folder. You can install the wheel using : pip install wheel/*.whl

Manually build the wheel

If you want to build the wheel yourself, you can use the build_and_setup_ultimatealvr.sh script. It will create a new directory tmp and build the wheel in there. It also includes the assets and libs folders needed when developing.

Copy necessary files/folders

Then you need to copy the assets and libs folders to the same directory as the script.

If you built the wheel in the previous step, you can copy the assets and libs folders from the tmp directory. If you used the already built wheel, you can find the 'assets' and 'libs' folders on the GitHub repository

The structure should look like this:

.
├── alpr_api.py
├── assets
│   ├── fonts
│   └── models
├── libs
│   ├── libxxxxxx.so
│   ├── ...
│   └── libxxxxxx.so
└── ...

Important notes

When building or developing the script, make sure to set the LD_LIBRARY_PATH environment variable to the libs folder (limitation of the ultimateALPR SDK).

export LD_LIBRARY_PATH=libs:$LD_LIBRARY_PATH