nuhmanpk / pyplatex

Simple , Scalable and Ready to use ANPR package for Automatic Number Plate Recognition

PyPlateX

High-Performance Scalable ANPR Package: Ready-to-Use, Simple, and Efficient License Plate Recognition

Unlock top-tier accuracy and scalability with cutting-edge ANPR solution in 3 line of code. Designed for seamless integration and ease of use, it delivers robust performance and reliability for all your license plate recognition needs.

Simple ready to use ANPR

Note: The ANPR.detect function is asynchronous, so ensure you use the await keyword when calling it within an async function.

Install from pypi.org

pip install pyplatex

from pyplatex import ANPR
anpr = ANPR()
det = await anpr.detect('./demo/plate-1.jpg')
print(det)

or

from pyplatex import ANPR
import asyncio

async def main():
    anpr = ANPR()
    plates = await anpr.detect('./demo/plate-1.jpg')
    print(plates)

# Run the async main function
asyncio.run(main())

the output would be like

{
    'is_plate': True, 
    'is_plate_confidence': 0.78, 
    'plate_number': 'MUN389', 
    'plate_number_confidence': 1.0
}

Args for anpr.detect()

Parameter	Default Value	Description
image_path	None	Path to the image file to be processed.
max_detections	1	Maximum number of license plates to detect in the image.
confidence	0.6	Confidence threshold for detecting a license plate. Only detections with confidence above this value will be considered.
save_image	False	If True, the detected plate image will be saved to disk.
padding	5	Padding around the detected license plate when saving the image.
folder_name	None	Directory name where the detected images will be saved. If save_image is True, this folder will be created if it does not exist.
use_ocr	True	If True, Optical Character Recognition (OCR) will be performed on the detected license plates.
return_tensor	False	If True, returns the image tensor of the detected license plates.
verbose	True	If True, logs detailed information during processing.

Dev TODO:

• Release a Inital Version
• Add a plate detection model
• Read and detect Plates
• Format output
• Integrate Cv2filters
• Change Cofidence to a round number
• Add a ocr Model
• Release a Initial Version
• Add a option to accept image as Tensor / numpy array
• Add auto filters tag

This is a pre-release version; there might be some bugs. If you encounter any issues or performance-related problems, please report them here. If you'd like to contribute to this project, you can create a pull request here.

Warning: Use this pre-release with caution as it may still have unresolved issues.

If you like this project, please consider giving it a star on Github! Your support is appreciated. If you want to contribute further, you can also sponsor the project through GitHub Sponsors. Every contribution helps improve and maintain the project for the community.

Happy Coding 🚀 ...