OCR 03: PaddleOCR

Catalog

• OCR 01: EasyOCR
• OCR 02: Tesseract-OCR
• OCR 03: PaddleOCR

About

• Git repository https://github.com/PaddlePaddle/PaddleOCR
• Online demo https://www.paddlepaddle.org.cn/hub/scene/ocr
• Installation Docs https://github.com/PaddlePaddle/PaddleOCR/blob/release/2.6/doc/doc_ch/quickstart.md

Install

python -m pip install paddlepaddle==2.3.2 -i https://pypi.tuna.tsinghua.edu.cn/simple
python -m pip install "paddleocr>=2.0.1"

If error ocurs during installation

 error: Microsoft Visual C++ 14.0 or greater is required. Get it with "Microsoft C++ Build Tools": https://visualstudio.microsoft.com/visual-cpp-build-tools/

Download buildTool Installer from https://visualstudio.microsoft.com/visual-cpp-build-tools/ and install the desktop c++ package, this will download several GiB files, taking a long time.

Usage

paddleocr --image_dir ./fp05b.jpg --use_angle_cls true --use_gpu false

If it is running for the first time, it will download the model files

Invoke in Python

from paddleocr import PaddleOCR, draw_ocr

# Paddleocr目前支持的多语言语种可以通过修改lang参数进行切换
# 例如`ch`, `en`, `fr`, `german`, `korean`, `japan`
ocr = PaddleOCR(use_angle_cls=True, lang="ch")  # need to run only once to download and load model into memory
img_path = './photos/fp04b.jpg'
result = ocr.ocr(img_path, cls=True)
for idx in range(len(result)):
    res = result[idx]
    for line in res:
        print(line)

# 显示结果
from PIL import Image
result = result[0]
image = Image.open(img_path).convert('RGB')
boxes = [line[0] for line in result]
txts = [line[1][0] for line in result]
scores = [line[1][1] for line in result]
im_show = draw_ocr(image, boxes, txts, scores, font_path='./msyh.ttc')
im_show = Image.fromarray(im_show)
im_show.save('fp04b_result.jpg')

Performance

• Run on CPU, it will take around 10 seconds to parsing a picture
• Correct rate is much better than Tesseract and EasyOCR, even when handling the cellphone photos.