pponnxcr 1.0

OCR based on onnxruntime with PaddleOCR models

pponnxcr - PaddlePaddle Onnxruntime OCR

OCR based on onnxruntime with PaddleOCR models
Refractor for ppocr-onnx which implements:

• Update rec model to v3
• Remove all unnecessary components, including draw-ocr
• Add En and Ja language support

Install

pip install pponnxcr

Usage

from pponnxcr import TextSystem
import cv2

ZH = TextSystem('zh')
JA = TextSystem('ja')
EN = TextSystem('en')

img = cv2.imread('test.png')

ZH.ocr_single_line(img)
ZH.ocr_lines([img, ...])
ZH.detect_and_ocr(img)

License

GNU AGPLv3

Which means that unless commercially licensed, any modification or use of this project in any way requires open source

Reference

• ppocr-onnx
• RapidOCR
• PaddleOCR