Mmocr usage guide

MMOCR It is a general visual framework OpenMMLab Optical character recognizer .

Install the configuration environment

MMOCR github Home page ：GitHub - open-mmlab/mmocr: OpenMMLab Text Detection, Recognition and Understanding Toolbox

pip install mmcv-full -f https://download.openmmlab.com/mmcv/dist/cu113/torch1.11.0/index.htmlpip install mmdet -i https://pypi.tuna.tsinghua.edu.cn/simplepip install lmdbpip install shapelypip install rapidfuzzpip install lanmspip install pyclipperpip install scikit-imagepip install imgaug

Verify whether the installation of the successful code

import torch, torchvisionimport mmcvfrom mmcv.ops import get_compiling_cuda_version, get_compiler_versionimport mmdetimport mmocrfrom mmocr.utils.ocr import MMOCRmmocr = MMOCR(det=None, recog='SAR', device='cpu')print('mmocr Loading succeeded ')

Text detection and text extraction

import torch, torchvisionimport mmcvfrom mmcv.ops import get_compiling_cuda_version, get_compiler_versionimport mmdetimport mmocrfrom mmocr.utils.ocr import MMOCR# mmocr = MMOCR(det=None, recog='SAR', device='cpu')# print('mmocr Loading succeeded ')if __name__ == '__main__':    detector = MMOCR(det='TextSnake', recog='SAR', device='cuda')    result = detector.readtext('demo/demo_densetext_det.jpg', output='output/demo_densetext_det.jpg')

Chinese detection and extraction

stay mmocr Create a new folder under the home directory /data/chineseocr/labels

Enter this folder to execute

wget http://download.openmmlab.com/mmocr/textrecog/sar/dict_printed_chinese_english_digits.txtwget http://download.openmmlab.com/mmocr/data/font.TTF

Download fonts and fonts

import torch, torchvisionimport mmcvfrom mmcv.ops import get_compiling_cuda_version, get_compiler_versionimport mmdetimport mmocrfrom mmocr.utils.ocr import MMOCR# mmocr = MMOCR(det=None, recog='SAR', device='cpu')# print('mmocr Loading succeeded ')if __name__ == '__main__':    detector = MMOCR(det='TextSnake', recog='SAR_CN', device='cuda')    result = detector.readtext('demo/demo_densetext_det.jpg', output='output/demo_densetext_det.jpg')