A robust CRNN-based OCR model for license plate recognition (Persian only)(V2). This model is a finetuned version of hezarai/crnn-fa-printed-96-long on the Persian license plate dataset (v1).

Note that this model only works on a cropped image of a plate, so you have to run an object(license plate) detection model and pass its output to this model to work.

Usage

pip install hezar
from hezar.models import Model

model = Model.load("hezarai/crnn-fa-printed-96-long")
plate_text = model.predict("assets/license_plate_ocr_example.jpg")
print(plate_text)  # Persian text of mixed numbers and characters might not show correctly in the console
['۵۷س۷۷۹۷۷']
Downloads last month
3,398
Inference Examples
Inference API (serverless) does not yet support hezar models for this pipeline type.

Model tree for hezarai/crnn-fa-license-plate-recognition-v2

Finetuned
(1)
this model

Dataset used to train hezarai/crnn-fa-license-plate-recognition-v2