Skip to content

P
PaddleOCR

s920243400 / PaddleOCR
与 Fork 源项目一致

Fork自 PaddlePaddle / PaddleOCR

通知 1
Star 1
Fork 0
P
PaddleOCR
“bbf15560a5a993057d948f16689cbd0e1e8540fa”上不存在“docs/zh/14-reference/_tcollector.mdx”
提交 684676f6 编写于 作者: W WenmuZhou
浏览文件
操作

init commit for build whl

上级 7c09c97d
显示空白变更内容
内联 并排
Showing with 814 addition and 12 deletion
.gitignore
浏览文件 @ 684676f6
......@@ -21,3 +21,7 @@ output/
*.log
.clang-format
.clang_format.hook
build/
dist/
paddleocr.egg-info/
\ No newline at end of file
MANIFEST.in 0 → 100644
浏览文件 @ 684676f6
include LICENSE.txt
include README.md
recursive-include ppocr/utils *.txt utility.py character.py check.py
recursive-include ppocr/data/det *.py
recursive-include ppocr/postprocess *.py
recursive-include ppocr/postprocess/lanms *.*
recursive-include tools/infer *.py
__init__.py 0 → 100644
浏览文件 @ 684676f6
# Copyright (c) 2020 PaddlePaddle Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
__all__ = ['PaddleOCR', 'draw_ocr']
from .paddleocr import PaddleOCR
from .tools.infer.utility import draw_ocr
doc/doc_ch/whl.md 0 → 100644
浏览文件 @ 684676f6
# paddleocr package使用说明
## 快速上手
### 安装whl包
pip安装
```bash
pip install paddleocr
```
本地构建并安装
```bash
python setup.py bdist_wheel
pip install dist/paddleocr-0.0.1-py3-none-any.whl
```
### 1. 代码使用
* 检测+识别全流程
```python
from paddleocr import PaddleOCR, draw_ocr
ocr = PaddleOCR(model_storage_directory='./model') # need to run only once to load model into memory
img_path = 'PaddleOCR/doc/imgs/11.jpg'
result = ocr.ocr(img_path)
for line in result:
print(line)
# 显示结果
from PIL import Image
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='/path/to/PaddleOCR/doc/simfang.ttf')
im_show = Image.fromarray(im_show)
im_show.save('result.jpg')
```
结果是一个list，每个item包含了文本框，文字和识别置信度
```bash
[[[24.0, 36.0], [304.0, 34.0], [304.0, 72.0], [24.0, 74.0]], ['纯臻营养护发素', 0.964739]]
[[[24.0, 80.0], [172.0, 80.0], [172.0, 104.0], [24.0, 104.0]], ['产品信息/参数', 0.98069626]]
[[[24.0, 109.0], [333.0, 109.0], [333.0, 136.0], [24.0, 136.0]], ['（45元/每公斤，100公斤起订）', 0.9676722]]
[[[22.0, 140.0], [284.0, 140.0], [284.0, 167.0], [22.0, 167.0]], ['每瓶22元，1000瓶起订）', 0.97444016]]
[[[22.0, 174.0], [85.0, 174.0], [85.0, 198.0], [22.0, 198.0]], ['【品牌】', 0.8187138]]
[[[89.0, 176.0], [301.0, 176.0], [301.0, 196.0], [89.0, 196.0]], ['：代加工方式/OEMODM', 0.9421848]]
[[[23.0, 205.0], [85.0, 205.0], [85.0, 229.0], [23.0, 229.0]], ['【品名】', 0.76008326]]
[[[88.0, 204.0], [235.0, 206.0], [235.0, 229.0], [88.0, 227.0]], ['：纯臻营养护发素', 0.9633639]]
[[[23.0, 236.0], [121.0, 236.0], [121.0, 261.0], [23.0, 261.0]], ['【产品编号】', 0.84101385]]
[[[110.0, 239.0], [239.0, 239.0], [239.0, 256.0], [110.0, 256.0]], ['1：YM-X-3011', 0.8621878]]
[[[414.0, 233.0], [430.0, 233.0], [430.0, 304.0], [414.0, 304.0]], ['ODM OEM', 0.9084018]]
[[[23.0, 268.0], [183.0, 268.0], [183.0, 292.0], [23.0, 292.0]], ['【净含量】：220ml', 0.9278281]]
[[[24.0, 301.0], [118.0, 301.0], [118.0, 321.0], [24.0, 321.0]], ['【适用人群】', 0.90901047]]
[[[127.0, 300.0], [254.0, 300.0], [254.0, 323.0], [127.0, 323.0]], ['：适合所有肤质', 0.95465785]]
[[[24.0, 332.0], [117.0, 332.0], [117.0, 353.0], [24.0, 353.0]], ['【主要成分】', 0.88936955]]
[[[139.0, 332.0], [236.0, 332.0], [236.0, 352.0], [139.0, 352.0]], ['鲸蜡硬脂醇', 0.9447544]]
[[[248.0, 332.0], [345.0, 332.0], [345.0, 352.0], [248.0, 352.0]], ['燕麦B-葡聚', 0.89748293]]
[[[54.0, 363.0], [232.0, 363.0], [232.0, 383.0], [54.0, 383.0]], [' 椰油酰胺丙基甜菜碱', 0.902023]]
[[[25.0, 364.0], [64.0, 364.0], [64.0, 383.0], [25.0, 383.0]], ['糖、', 0.985203]]
[[[244.0, 363.0], [281.0, 363.0], [281.0, 382.0], [244.0, 382.0]], ['泛服', 0.44537082]]
[[[367.0, 367.0], [475.0, 367.0], [475.0, 388.0], [367.0, 388.0]], ['（成品包材）', 0.9834532]]
[[[24.0, 395.0], [120.0, 395.0], [120.0, 416.0], [24.0, 416.0]], ['【主要功能】', 0.88684446]]
[[[128.0, 397.0], [273.0, 397.0], [273.0, 414.0], [128.0, 414.0]], ['：可紧致头发磷层', 0.9342501]]
[[[265.0, 395.0], [361.0, 395.0], [361.0, 415.0], [265.0, 415.0]], ['琴，从而达到', 0.8253762]]
[[[25.0, 425.0], [372.0, 425.0], [372.0, 448.0], [25.0, 448.0]], ['即时持久改善头发光泽的效果，给干燥的头', 0.97785276]]
[[[26.0, 457.0], [137.0, 457.0], [137.0, 477.0], [26.0, 477.0]], ['发足够的滋养', 0.9577897]]
```
结果可视化
<div align="center">
<img src="../imgs_results/whl/11_det_rec.jpg" width="800">
</div>
* 单独执行检测
```python
from paddleocr import PaddleOCR, draw_ocr
ocr = PaddleOCR(model_storage_directory='./model') # need to run only once to load model into memory
img_path = 'PaddleOCR/doc/imgs/11.jpg'
result = ocr.ocr(img_path,rec=False)
for line in result:
print(line)
# 显示结果
from PIL import Image
image = Image.open(img_path).convert('RGB')
im_show = draw_ocr(image, result, txts=None, scores=None, font_path='/path/to/PaddleOCR/doc/simfang.ttf')
im_show = Image.fromarray(im_show)
im_show.save('result.jpg')
```
结果是一个list，每个item只包含文本框
```bash
[[26.0, 457.0], [137.0, 457.0], [137.0, 477.0], [26.0, 477.0]]
[[25.0, 425.0], [372.0, 425.0], [372.0, 448.0], [25.0, 448.0]]
[[128.0, 397.0], [273.0, 397.0], [273.0, 414.0], [128.0, 414.0]]
[[265.0, 395.0], [361.0, 395.0], [361.0, 415.0], [265.0, 415.0]]
[[24.0, 395.0], [120.0, 395.0], [120.0, 416.0], [24.0, 416.0]]
[[367.0, 367.0], [475.0, 367.0], [475.0, 388.0], [367.0, 388.0]]
[[54.0, 363.0], [232.0, 363.0], [232.0, 383.0], [54.0, 383.0]]
[[25.0, 364.0], [64.0, 364.0], [64.0, 383.0], [25.0, 383.0]]
[[244.0, 363.0], [281.0, 363.0], [281.0, 382.0], [244.0, 382.0]]
[[248.0, 332.0], [345.0, 332.0], [345.0, 352.0], [248.0, 352.0]]
[[139.0, 332.0], [236.0, 332.0], [236.0, 352.0], [139.0, 352.0]]
[[24.0, 332.0], [117.0, 332.0], [117.0, 353.0], [24.0, 353.0]]
[[127.0, 300.0], [254.0, 300.0], [254.0, 323.0], [127.0, 323.0]]
[[24.0, 301.0], [118.0, 301.0], [118.0, 321.0], [24.0, 321.0]]
[[23.0, 268.0], [183.0, 268.0], [183.0, 292.0], [23.0, 292.0]]
[[110.0, 239.0], [239.0, 239.0], [239.0, 256.0], [110.0, 256.0]]
[[23.0, 236.0], [121.0, 236.0], [121.0, 261.0], [23.0, 261.0]]
[[414.0, 233.0], [430.0, 233.0], [430.0, 304.0], [414.0, 304.0]]
[[88.0, 204.0], [235.0, 206.0], [235.0, 229.0], [88.0, 227.0]]
[[23.0, 205.0], [85.0, 205.0], [85.0, 229.0], [23.0, 229.0]]
[[89.0, 176.0], [301.0, 176.0], [301.0, 196.0], [89.0, 196.0]]
[[22.0, 174.0], [85.0, 174.0], [85.0, 198.0], [22.0, 198.0]]
[[22.0, 140.0], [284.0, 140.0], [284.0, 167.0], [22.0, 167.0]]
[[24.0, 109.0], [333.0, 109.0], [333.0, 136.0], [24.0, 136.0]]
[[24.0, 80.0], [172.0, 80.0], [172.0, 104.0], [24.0, 104.0]]
[[24.0, 36.0], [304.0, 34.0], [304.0, 72.0], [24.0, 74.0]]
```
结果可视化
<div align="center">
<img src="../imgs_results/whl/11_det.jpg" width="800">
</div>
* 单独执行识别
```python
from paddleocr import PaddleOCR
ocr = PaddleOCR(model_storage_directory='./model') # need to run only once to load model into memory
img_path = 'PaddleOCR/doc/imgs_words/ch/word_1.jpg'
result = ocr.ocr(img_path,det=False)
for line in result:
print(line)
```
结果是一个list，每个item只包含识别结果和识别置信度
```bash
['韩国小馆', 0.9907421]
```
### 通过命令行使用
查看帮助信息
```bash
paddleocr -h
```
* 检测+识别全流程
```bash
paddleocr --image_dir PaddleOCR/doc/imgs/11.jpg
```
结果是一个list，每个item包含了文本框，文字和识别置信度
```bash
[[[24.0, 36.0], [304.0, 34.0], [304.0, 72.0], [24.0, 74.0]], ['纯臻营养护发素', 0.964739]]
[[[24.0, 80.0], [172.0, 80.0], [172.0, 104.0], [24.0, 104.0]], ['产品信息/参数', 0.98069626]]
[[[24.0, 109.0], [333.0, 109.0], [333.0, 136.0], [24.0, 136.0]], ['（45元/每公斤，100公斤起订）', 0.9676722]]
[[[22.0, 140.0], [284.0, 140.0], [284.0, 167.0], [22.0, 167.0]], ['每瓶22元，1000瓶起订）', 0.97444016]]
[[[22.0, 174.0], [85.0, 174.0], [85.0, 198.0], [22.0, 198.0]], ['【品牌】', 0.8187138]]
[[[89.0, 176.0], [301.0, 176.0], [301.0, 196.0], [89.0, 196.0]], ['：代加工方式/OEMODM', 0.9421848]]
[[[23.0, 205.0], [85.0, 205.0], [85.0, 229.0], [23.0, 229.0]], ['【品名】', 0.76008326]]
[[[88.0, 204.0], [235.0, 206.0], [235.0, 229.0], [88.0, 227.0]], ['：纯臻营养护发素', 0.9633639]]
[[[23.0, 236.0], [121.0, 236.0], [121.0, 261.0], [23.0, 261.0]], ['【产品编号】', 0.84101385]]
[[[110.0, 239.0], [239.0, 239.0], [239.0, 256.0], [110.0, 256.0]], ['1：YM-X-3011', 0.8621878]]
[[[414.0, 233.0], [430.0, 233.0], [430.0, 304.0], [414.0, 304.0]], ['ODM OEM', 0.9084018]]
[[[23.0, 268.0], [183.0, 268.0], [183.0, 292.0], [23.0, 292.0]], ['【净含量】：220ml', 0.9278281]]
[[[24.0, 301.0], [118.0, 301.0], [118.0, 321.0], [24.0, 321.0]], ['【适用人群】', 0.90901047]]
[[[127.0, 300.0], [254.0, 300.0], [254.0, 323.0], [127.0, 323.0]], ['：适合所有肤质', 0.95465785]]
[[[24.0, 332.0], [117.0, 332.0], [117.0, 353.0], [24.0, 353.0]], ['【主要成分】', 0.88936955]]
[[[139.0, 332.0], [236.0, 332.0], [236.0, 352.0], [139.0, 352.0]], ['鲸蜡硬脂醇', 0.9447544]]
[[[248.0, 332.0], [345.0, 332.0], [345.0, 352.0], [248.0, 352.0]], ['燕麦B-葡聚', 0.89748293]]
[[[54.0, 363.0], [232.0, 363.0], [232.0, 383.0], [54.0, 383.0]], [' 椰油酰胺丙基甜菜碱', 0.902023]]
[[[25.0, 364.0], [64.0, 364.0], [64.0, 383.0], [25.0, 383.0]], ['糖、', 0.985203]]
[[[244.0, 363.0], [281.0, 363.0], [281.0, 382.0], [244.0, 382.0]], ['泛服', 0.44537082]]
[[[367.0, 367.0], [475.0, 367.0], [475.0, 388.0], [367.0, 388.0]], ['（成品包材）', 0.9834532]]
[[[24.0, 395.0], [120.0, 395.0], [120.0, 416.0], [24.0, 416.0]], ['【主要功能】', 0.88684446]]
[[[128.0, 397.0], [273.0, 397.0], [273.0, 414.0], [128.0, 414.0]], ['：可紧致头发磷层', 0.9342501]]
[[[265.0, 395.0], [361.0, 395.0], [361.0, 415.0], [265.0, 415.0]], ['琴，从而达到', 0.8253762]]
[[[25.0, 425.0], [372.0, 425.0], [372.0, 448.0], [25.0, 448.0]], ['即时持久改善头发光泽的效果，给干燥的头', 0.97785276]]
[[[26.0, 457.0], [137.0, 457.0], [137.0, 477.0], [26.0, 477.0]], ['发足够的滋养', 0.9577897]]
```
* 单独执行检测
```bash
paddleocr --image_dir PaddleOCR/doc/imgs/11.jpg --rec false
```
结果是一个list，每个item只包含文本框
```bash
[[26.0, 457.0], [137.0, 457.0], [137.0, 477.0], [26.0, 477.0]]
[[25.0, 425.0], [372.0, 425.0], [372.0, 448.0], [25.0, 448.0]]
[[128.0, 397.0], [273.0, 397.0], [273.0, 414.0], [128.0, 414.0]]
[[265.0, 395.0], [361.0, 395.0], [361.0, 415.0], [265.0, 415.0]]
[[24.0, 395.0], [120.0, 395.0], [120.0, 416.0], [24.0, 416.0]]
[[367.0, 367.0], [475.0, 367.0], [475.0, 388.0], [367.0, 388.0]]
[[54.0, 363.0], [232.0, 363.0], [232.0, 383.0], [54.0, 383.0]]
[[25.0, 364.0], [64.0, 364.0], [64.0, 383.0], [25.0, 383.0]]
[[244.0, 363.0], [281.0, 363.0], [281.0, 382.0], [244.0, 382.0]]
[[248.0, 332.0], [345.0, 332.0], [345.0, 352.0], [248.0, 352.0]]
[[139.0, 332.0], [236.0, 332.0], [236.0, 352.0], [139.0, 352.0]]
[[24.0, 332.0], [117.0, 332.0], [117.0, 353.0], [24.0, 353.0]]
[[127.0, 300.0], [254.0, 300.0], [254.0, 323.0], [127.0, 323.0]]
[[24.0, 301.0], [118.0, 301.0], [118.0, 321.0], [24.0, 321.0]]
[[23.0, 268.0], [183.0, 268.0], [183.0, 292.0], [23.0, 292.0]]
[[110.0, 239.0], [239.0, 239.0], [239.0, 256.0], [110.0, 256.0]]
[[23.0, 236.0], [121.0, 236.0], [121.0, 261.0], [23.0, 261.0]]
[[414.0, 233.0], [430.0, 233.0], [430.0, 304.0], [414.0, 304.0]]
[[88.0, 204.0], [235.0, 206.0], [235.0, 229.0], [88.0, 227.0]]
[[23.0, 205.0], [85.0, 205.0], [85.0, 229.0], [23.0, 229.0]]
[[89.0, 176.0], [301.0, 176.0], [301.0, 196.0], [89.0, 196.0]]
[[22.0, 174.0], [85.0, 174.0], [85.0, 198.0], [22.0, 198.0]]
[[22.0, 140.0], [284.0, 140.0], [284.0, 167.0], [22.0, 167.0]]
[[24.0, 109.0], [333.0, 109.0], [333.0, 136.0], [24.0, 136.0]]
[[24.0, 80.0], [172.0, 80.0], [172.0, 104.0], [24.0, 104.0]]
[[24.0, 36.0], [304.0, 34.0], [304.0, 72.0], [24.0, 74.0]]
```
* 单独执行识别
```bash
paddleocr --image_dir PaddleOCR/doc/imgs_words/ch/word_1.jpg --det false
```
结果是一个list，每个item只包含识别结果和识别置信度
```bash
['韩国小馆', 0.9907421]
```
## 参数说明
| 字段 | 说明 | 默认值 |
|-------------------------|----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|-------------------------|
| use_gpu | 是否使用GPU | TRUE |
| gpu_mem | 初始化占用的GPU内存大小 | 8000M |
| image_dir | 通过命令行调用时执行预测的图片或文件夹路径 | |
| det_algorithm | 使用的检测算法类型 | DB |
| det_model_name | 有两种使用方式： 1. 检测算法名称，此名称必须在支持列表内(目前只内置了ch_det_mv3_db)，传入错误参数时会显示支持的列表 2. 自己转换好的inference模型路径，此时模型路径下必须包含model和params文件。选择此方式时，需要手动指定det_algorithm的值 | ch_det_mv3_db |
| det_max_side_len | 检测算法前向时图片长边的最大尺寸，当长边超出这个值时会将长边resize到这个大小，短边等比例缩放 | 960 |
| det_db_thresh | DB模型输出预测图的二值化阈值 | 0.3 |
| det_db_box_thresh | DB模型输出框的阈值，低于此值的预测框会被丢弃 | 0.5 |
| det_db_unclip_ratio | DB模型输出框扩大的比例 | 2 |
| det_east_score_thresh | EAST模型输出预测图的二值化阈值 | 0.8 |
| det_east_cover_thresh | EAST模型输出框的阈值，低于此值的预测框会被丢弃 | 0.1 |
| det_east_nms_thresh | EAST模型输出框NMS的阈值 | 0.2 |
| rec_algorithm | 使用的识别算法类型 | CRNN |
| rec_model_name | 有两种使用方式： 1. 识别算法名称，此名称必须在支持列表内(目前支持CRNN,Rosetta,STAR,RARE等算法,但是内置的只有ch_rec_mv3_crnn_enhance)，传入错误参数时会显示支持的列表 2. 自己转换好的inference模型路径，此时模型路径下必须包含model和params文件。选择此方式时，需要手动指定rec_algorithm的值 | ch_rec_mv3_crnn_enhance |
| rec_image_shape | 识别算法的输入图片尺寸 | "3,32,320" |
| rec_char_type | 识别算法的字符类型，中文(ch)或英文(en) | ch |
| rec_batch_num | 进行识别时，同时前向的图片数 | 30 |
| rec_char_dict_path | 识别模型字典路径，当rec_model_name使用方式2传参时需要修改为自己的路径 | |
| use_space_char | 是否识别空格 | TRUE |
| enable_mkldnn | 是否启用mkldnn | FALSE |
| model_storage_directory | 下载模型保存路径 | ～/.paddleocr |
| det | 前向时使用启动检测 | TRUE |
| rec | 前向时是否启动识别 | TRUE |
doc/doc_en/whl.md 0 → 100644
浏览文件 @ 684676f6
# paddleocr package
## Get started quickly
### install package
install by pypi
```bash
pip install paddleocr
```
build own whl package and install
```bash
python setup.py bdist_wheel
pip install dist/paddleocr-0.0.1-py3-none-any.whl
```
### 1. Use by code
* detection and recognition
```python
from paddleocr import PaddleOCR,draw_ocr
ocr = PaddleOCR(model_storage_directory='./model') # need to run only once to load model into memory
img_path = 'PaddleOCR/doc/imgs_en/img_12.jpg'
result = ocr.ocr(img_path)
for line in result:
print(line)
# draw result
from PIL import Image
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='/path/to/PaddleOCR/doc/simfang.ttf')
im_show = Image.fromarray(im_show)
im_show.save('result.jpg')
```
Output will be a list, each item contains bounding box, text and recognition confidence
```bash
[[[442.0, 173.0], [1169.0, 173.0], [1169.0, 225.0], [442.0, 225.0]], ['ACKNOWLEDGEMENTS', 0.99283075]]
[[[393.0, 340.0], [1207.0, 342.0], [1207.0, 389.0], [393.0, 387.0]], ['We would like to thank all the designers and', 0.9357758]]
[[[399.0, 398.0], [1204.0, 398.0], [1204.0, 433.0], [399.0, 433.0]], ['contributors whohave been involved in the', 0.9592447]]
[[[395.0, 443.0], [1211.0, 443.0], [1211.0, 489.0], [395.0, 489.0]], ['production of this book;their contributions', 0.9713175]]
[[[395.0, 497.0], [1209.0, 495.0], [1209.0, 531.0], [395.0, 533.0]], ['have been indispensable to its creation.We', 0.96009934]]
[[[393.0, 545.0], [1212.0, 545.0], [1212.0, 591.0], [393.0, 591.0]], ['would also like to express our gratitude to al', 0.9371007]]
[[[393.0, 595.0], [1212.0, 593.0], [1212.0, 635.0], [393.0, 637.0]], ['the producers for their invaluable opinions', 0.96872145]]
[[[393.0, 645.0], [1209.0, 645.0], [1209.0, 685.0], [393.0, 685.0]], ['and assistance throughout this proiect.Andto', 0.94448787]]
[[[392.0, 697.0], [1212.0, 693.0], [1212.0, 735.0], [392.0, 739.0]], ['the many others whose names are not credited', 0.93633145]]
[[[397.0, 753.0], [689.0, 755.0], [689.0, 786.0], [397.0, 784.0]], ['buthavemades', 0.99324507]]
[[[813.0, 749.0], [1212.0, 747.0], [1212.0, 784.0], [813.0, 786.0]], ['inputin this book, we', 0.9166398]]
[[[675.0, 760.0], [799.0, 755.0], [799.0, 778.0], [675.0, 784.0]], ['speciti', 0.9063535]]
[[[393.0, 801.0], [715.0, 805.0], [715.0, 839.0], [393.0, 836.0]], ['thankyouforyoul', 0.92475533]]
[[[756.0, 812.0], [805.0, 812.0], [805.0, 830.0], [756.0, 830.0]], ['P', 0.14887337]]
[[[820.0, 803.0], [1085.0, 801.0], [1085.0, 836.0], [820.0, 838.0]], ['nuoussupport', 0.9898951]]
```
Visualization of results
<div align="center">
<img src="../imgs_results/whl/12_det_rec.jpg" width="800">
</div>
* only detection
```python
from paddleocr import PaddleOCR,draw_ocr
ocr = PaddleOCR(model_storage_directory='./model') # need to run only once to load model into memory
img_path = 'PaddleOCR/doc/imgs_en/img_12.jpg'
result = ocr.ocr(img_path,rec=False)
for line in result:
print(line)
# draw result
from PIL import Image
image = Image.open(img_path).convert('RGB')
im_show = draw_ocr(image, result, txts=None, scores=None, font_path='/path/to/PaddleOCR/doc/simfang.ttf')
im_show = Image.fromarray(im_show)
im_show.save('result.jpg')
```
Output will be a list, each item only contains bounding box
```bash
[[756.0, 812.0], [805.0, 812.0], [805.0, 830.0], [756.0, 830.0]]
[[820.0, 803.0], [1085.0, 801.0], [1085.0, 836.0], [820.0, 838.0]]
[[393.0, 801.0], [715.0, 805.0], [715.0, 839.0], [393.0, 836.0]]
[[675.0, 760.0], [799.0, 755.0], [799.0, 778.0], [675.0, 784.0]]
[[397.0, 753.0], [689.0, 755.0], [689.0, 786.0], [397.0, 784.0]]
[[813.0, 749.0], [1212.0, 747.0], [1212.0, 784.0], [813.0, 786.0]]
[[392.0, 697.0], [1212.0, 693.0], [1212.0, 735.0], [392.0, 739.0]]
[[393.0, 645.0], [1209.0, 645.0], [1209.0, 685.0], [393.0, 685.0]]
[[393.0, 595.0], [1212.0, 593.0], [1212.0, 635.0], [393.0, 637.0]]
[[393.0, 545.0], [1212.0, 545.0], [1212.0, 591.0], [393.0, 591.0]]
[[395.0, 497.0], [1209.0, 495.0], [1209.0, 531.0], [395.0, 533.0]]
[[395.0, 443.0], [1211.0, 443.0], [1211.0, 489.0], [395.0, 489.0]]
[[399.0, 398.0], [1204.0, 398.0], [1204.0, 433.0], [399.0, 433.0]]
[[393.0, 340.0], [1207.0, 342.0], [1207.0, 389.0], [393.0, 387.0]]
[[442.0, 173.0], [1169.0, 173.0], [1169.0, 225.0], [442.0, 225.0]]
```
Visualization of results
<div align="center">
<img src="../imgs_results/whl/12_det.jpg" width="800">
</div>
* only recognition
```python
from paddleocr import PaddleOCR
ocr = PaddleOCR(model_storage_directory='./model') # need to run only once to load model into memory
img_path = 'PaddleOCR/doc/imgs_words_en/word_10.png'
result = ocr.ocr(img_path,det=False)
for line in result:
print(line)
```
Output will be a list, each item contains text and recognition confidence
```bash
['PAIN', 0.990372]
```
### Use by command line
show help information
```bash
paddleocr -h
```
* detection and recognition
```bash
paddleocr --image_dir PaddleOCR/doc/imgs_en/img_12.jpg
```
Output will be a list, each item contains bounding box, text and recognition confidence
```bash
[[[442.0, 173.0], [1169.0, 173.0], [1169.0, 225.0], [442.0, 225.0]], ['ACKNOWLEDGEMENTS', 0.99283075]]
[[[393.0, 340.0], [1207.0, 342.0], [1207.0, 389.0], [393.0, 387.0]], ['We would like to thank all the designers and', 0.9357758]]
[[[399.0, 398.0], [1204.0, 398.0], [1204.0, 433.0], [399.0, 433.0]], ['contributors whohave been involved in the', 0.9592447]]
[[[395.0, 443.0], [1211.0, 443.0], [1211.0, 489.0], [395.0, 489.0]], ['production of this book;their contributions', 0.9713175]]
[[[395.0, 497.0], [1209.0, 495.0], [1209.0, 531.0], [395.0, 533.0]], ['have been indispensable to its creation.We', 0.96009934]]
[[[393.0, 545.0], [1212.0, 545.0], [1212.0, 591.0], [393.0, 591.0]], ['would also like to express our gratitude to al', 0.9371007]]
[[[393.0, 595.0], [1212.0, 593.0], [1212.0, 635.0], [393.0, 637.0]], ['the producers for their invaluable opinions', 0.96872145]]
[[[393.0, 645.0], [1209.0, 645.0], [1209.0, 685.0], [393.0, 685.0]], ['and assistance throughout this proiect.Andto', 0.94448787]]
[[[392.0, 697.0], [1212.0, 693.0], [1212.0, 735.0], [392.0, 739.0]], ['the many others whose names are not credited', 0.93633145]]
[[[397.0, 753.0], [689.0, 755.0], [689.0, 786.0], [397.0, 784.0]], ['buthavemades', 0.99324507]]
[[[813.0, 749.0], [1212.0, 747.0], [1212.0, 784.0], [813.0, 786.0]], ['inputin this book, we', 0.9166398]]
[[[675.0, 760.0], [799.0, 755.0], [799.0, 778.0], [675.0, 784.0]], ['speciti', 0.9063535]]
[[[393.0, 801.0], [715.0, 805.0], [715.0, 839.0], [393.0, 836.0]], ['thankyouforyoul', 0.92475533]]
[[[756.0, 812.0], [805.0, 812.0], [805.0, 830.0], [756.0, 830.0]], ['P', 0.14887337]]
[[[820.0, 803.0], [1085.0, 801.0], [1085.0, 836.0], [820.0, 838.0]], ['nuoussupport', 0.9898951]]
```
* only detection
```bash
paddleocr --image_dir PaddleOCR/doc/imgs_en/img_12.jpg --rec false
```
Output will be a list, each item only contains bounding box
```bash
[[756.0, 812.0], [805.0, 812.0], [805.0, 830.0], [756.0, 830.0]]
[[820.0, 803.0], [1085.0, 801.0], [1085.0, 836.0], [820.0, 838.0]]
[[393.0, 801.0], [715.0, 805.0], [715.0, 839.0], [393.0, 836.0]]
[[675.0, 760.0], [799.0, 755.0], [799.0, 778.0], [675.0, 784.0]]
[[397.0, 753.0], [689.0, 755.0], [689.0, 786.0], [397.0, 784.0]]
[[813.0, 749.0], [1212.0, 747.0], [1212.0, 784.0], [813.0, 786.0]]
[[392.0, 697.0], [1212.0, 693.0], [1212.0, 735.0], [392.0, 739.0]]
[[393.0, 645.0], [1209.0, 645.0], [1209.0, 685.0], [393.0, 685.0]]
[[393.0, 595.0], [1212.0, 593.0], [1212.0, 635.0], [393.0, 637.0]]
[[393.0, 545.0], [1212.0, 545.0], [1212.0, 591.0], [393.0, 591.0]]
[[395.0, 497.0], [1209.0, 495.0], [1209.0, 531.0], [395.0, 533.0]]
[[395.0, 443.0], [1211.0, 443.0], [1211.0, 489.0], [395.0, 489.0]]
[[399.0, 398.0], [1204.0, 398.0], [1204.0, 433.0], [399.0, 433.0]]
[[393.0, 340.0], [1207.0, 342.0], [1207.0, 389.0], [393.0, 387.0]]
[[442.0, 173.0], [1169.0, 173.0], [1169.0, 225.0], [442.0, 225.0]]
```
* only recognition
```bash
paddleocr --image_dir PaddleOCR/doc/imgs_words_en/word_10.png --det false
```
Output will be a list, each item contains text and recognition confidence
```bash
['PAIN', 0.990372]
```
## Parameter Description
| Parameter | Description | Default value |
|-------------------------|----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|-------------------------|
| use_gpu | use GPU or not | TRUE |
| gpu_mem | GPU memory size used for initialization | 8000M |
| image_dir | The images path or folder path for predicting when used by the command line | |
| det_algorithm | Type of detection algorithm selected | DB |
| det_model_name | There are two ways to use: 1. The name of the detection algorithm which must be in the support list(only ch_det_mv3_db is built in currently), and the supported list will be displayed when the wrong parameter is passed in. 2. The path of the inference model that has been converted by yourself. At this time, the model path must contains model and params files. When choosing this method, you need to give the name of det_algorithm | ch_det_mv3_db |
| det_max_side_len | The maximum size of the long side of the image. When the long side exceeds this value, the long side will be resized to this size, and the short side will be scaled proportionally | 960 |
| det_db_thresh | Binarization threshold value of DB output map | 0.3 |
| det_db_box_thresh | The threshold value of the DB output box. Boxes score lower than this value will be discarded | 0.5 |
| det_db_unclip_ratio | The expanded ratio of DB output box | 2 |
| det_east_score_thresh | Binarization threshold value of EAST output map | 0.8 |
| det_east_cover_thresh | The threshold value of the EAST output box. Boxes score lower than this value will be discarded | 0.1 |
| det_east_nms_thresh | The NMS threshold value of EAST model output box | 0.2 |
| rec_algorithm | Type of recognition algorithm selected | CRNN |
| rec_model_name | There are two ways to use: 1. The name of the recognition algorithm which must be in the support list(only supports CRNN, Rosetta, STAR, RARE and other algorithms currently, but only ch_rec_mv3_crnn_enhance is built-in), and the supported list will be displayed when the wrong parameter is passed in. 2. The path of the inference model that has been converted by yourself. At this time, the model path must contains model and params files. When choosing this method, you need to give the name of rec_algorithm | ch_rec_mv3_crnn_enhance |
| rec_image_shape | image shape of recognition algorithm | "3,32,320" |
| rec_char_type | Character type of recognition algorithm, Chinese (ch) or English (en) | ch |
| rec_batch_num | When performing recognition, the batchsize of forward images | 30 |
| rec_char_dict_path | the alphabet path which needs to be modified to your own path when `rec_model_Name` use mode 2 | |
| use_space_char | Whether to recognize spaces | TRUE |
| enable_mkldnn | Whether to enable mkldnn | FALSE |
| model_storage_directory | Download model save path when det_model_name or rec_model_name use mode 1 | ～/.paddleocr |
| det | Enable detction when `ppocr.ocr` func exec | TRUE |
| rec | Enable detction when `ppocr.ocr` func exec | TRUE |
paddleocr.py 0 → 100644
浏览文件 @ 684676f6
# Copyright (c) 2020 PaddlePaddle Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
import os
import sys
__dir__ = os.path.dirname(__file__)
sys.path.append(os.path.join(__dir__, ''))
import cv2
import numpy as np
from pathlib import Path
import tarfile
import requests
from tqdm import tqdm
from tools.infer import predict_system
from ppocr.utils.utility import initial_logger
logger = initial_logger()
from ppocr.utils.utility import check_and_read_gif
__all__ = ['PaddleOCR']
model_params = {
'ch_det_mv3_db': {
'url':
'https://paddleocr.bj.bcebos.com/ch_models/ch_det_mv3_db_infer.tar',
'algorithm': 'DB',
},
'ch_rec_mv3_crnn_enhance': {
'url':
'https://paddleocr.bj.bcebos.com/ch_models/ch_rec_mv3_crnn_enhance_infer.tar',
'algorithm': 'CRNN'
},
}
SUPPORT_DET_MODEL = ['DB']
SUPPORT_REC_MODEL = ['Rosetta', 'CRNN', 'STARNet', 'RARE']
def download_with_progressbar(url, save_path):
response = requests.get(url, stream=True)
total_size_in_bytes = int(response.headers.get('content-length', 0))
block_size = 1024 # 1 Kibibyte
progress_bar = tqdm(total=total_size_in_bytes, unit='iB', unit_scale=True)
with open(save_path, 'wb') as file:
for data in response.iter_content(block_size):
progress_bar.update(len(data))
file.write(data)
progress_bar.close()
if total_size_in_bytes != 0 and progress_bar.n != total_size_in_bytes:
logger.error("ERROR, something went wrong")
sys.exit(0)
def download_and_unzip(url, model_storage_directory):
tmp_path = os.path.join(model_storage_directory, url.split('/')[-1])
print('download {} to {}'.format(url, tmp_path))
os.makedirs(model_storage_directory, exist_ok=True)
download_with_progressbar(url, tmp_path)
with tarfile.open(tmp_path, 'r') as tarObj:
for filename in tarObj.getnames():
tarObj.extract(filename, model_storage_directory)
os.remove(tmp_path)
def maybe_download(model_storage_directory, model_name, mode='det'):
algorithm = None
# using custom model
if os.path.exists(os.path.join(model_name, 'model')) and os.path.exists(
os.path.join(model_name, 'params')):
return model_name, algorithm
# using the model of paddleocr
model_path = os.path.join(model_storage_directory, model_name)
if not os.path.exists(os.path.join(model_path,
'model')) or not os.path.exists(
os.path.join(model_path, 'params')):
assert model_name in model_params, 'model must in {}'.format(
model_params.keys())
download_and_unzip(model_params[model_name]['url'],
model_storage_directory)
algorithm = model_params[model_name]['algorithm']
return model_path, algorithm
def parse_args():
import argparse
def str2bool(v):
return v.lower() in ("true", "t", "1")
parser = argparse.ArgumentParser()
# params for prediction engine
parser.add_argument("--use_gpu", type=str2bool, default=True)
parser.add_argument("--ir_optim", type=str2bool, default=True)
parser.add_argument("--use_tensorrt", type=str2bool, default=False)
parser.add_argument("--gpu_mem", type=int, default=8000)
# params for text detector
parser.add_argument("--image_dir", type=str)
parser.add_argument("--det_algorithm", type=str, default='DB')
parser.add_argument("--det_model_name", type=str, default='ch_det_mv3_db')
parser.add_argument("--det_max_side_len", type=float, default=960)
# DB parmas
parser.add_argument("--det_db_thresh", type=float, default=0.3)
parser.add_argument("--det_db_box_thresh", type=float, default=0.5)
parser.add_argument("--det_db_unclip_ratio", type=float, default=2.0)
# EAST parmas
parser.add_argument("--det_east_score_thresh", type=float, default=0.8)
parser.add_argument("--det_east_cover_thresh", type=float, default=0.1)
parser.add_argument("--det_east_nms_thresh", type=float, default=0.2)
# params for text recognizer
parser.add_argument("--rec_algorithm", type=str, default='CRNN')
parser.add_argument(
"--rec_model_name", type=str, default='ch_rec_mv3_crnn_enhance')
parser.add_argument("--rec_image_shape", type=str, default="3, 32, 320")
parser.add_argument("--rec_char_type", type=str, default='ch')
parser.add_argument("--rec_batch_num", type=int, default=30)
parser.add_argument(
"--rec_char_dict_path",
type=str,
default="./ppocr/utils/ppocr_keys_v1.txt")
parser.add_argument("--use_space_char", type=bool, default=True)
parser.add_argument("--enable_mkldnn", type=bool, default=False)
parser.add_argument("--model_storage_directory", type=str, default=False)
parser.add_argument("--det", type=str2bool, default=True)
parser.add_argument("--rec", type=str2bool, default=True)
return parser.parse_args()
class PaddleOCR(predict_system.TextSystem):
def __init__(self,
det_model_name='ch_det_mv3_db',
rec_model_name='ch_rec_mv3_crnn_enhance',
model_storage_directory=None,
log_level=20,
**kwargs):
"""
paddleocr package
args:
det_model_name: det_model name, keep same with filename in paddleocr. default is ch_det_mv3_db
det_model_name: rec_model name, keep same with filename in paddleocr. default is ch_rec_mv3_crnn_enhance
model_storage_directory: model save path. default is ~/.paddleocr
det model will save to model_storage_directory/det_model
rec model will save to model_storage_directory/rec_model
log_level:
**kwargs: other params show in paddleocr --help
"""
logger.setLevel(log_level)
postprocess_params = parse_args()
# init model dir
if model_storage_directory:
self.model_storage_directory = model_storage_directory
else:
self.model_storage_directory = os.path.expanduser(
"~/.paddleocr/") + '/model'
Path(self.model_storage_directory).mkdir(parents=True, exist_ok=True)
# download model
det_model_path, det_algorithm = maybe_download(
self.model_storage_directory, det_model_name, 'det')
rec_model_path, rec_algorithm = maybe_download(
self.model_storage_directory, rec_model_name, 'rec')
# update model and post_process params
postprocess_params.__dict__.update(**kwargs)
postprocess_params.det_model_dir = det_model_path
postprocess_params.rec_model_dir = rec_model_path
if det_algorithm is not None:
postprocess_params.det_algorithm = det_algorithm
if rec_algorithm is not None:
postprocess_params.rec_algorithm = rec_algorithm
if postprocess_params.det_algorithm not in SUPPORT_DET_MODEL:
logger.error('det_algorithm must in {}'.format(SUPPORT_DET_MODEL))
sys.exit(0)
if postprocess_params.rec_algorithm not in SUPPORT_REC_MODEL:
logger.error('rec_algorithm must in {}'.format(SUPPORT_REC_MODEL))
sys.exit(0)
postprocess_params.rec_char_dict_path = Path(
__file__).parent / postprocess_params.rec_char_dict_path
# init det_model and rec_model
super().__init__(postprocess_params)
def ocr(self, img, det=True, rec=True):
"""
ocr with paddleocr
args：
img: img for ocr, support ndarray, img_path and list or ndarray
det: use text detection or not, if false, only rec will be exec. default is True
rec: use text recognition or not, if false, only det will be exec. default is True
"""
assert isinstance(img, (np.ndarray, list, str))
if isinstance(img, str):
image_file = img
img, flag = check_and_read_gif(image_file)
if not flag:
img = cv2.imread(image_file)
if img is None:
logger.error("error in loading image:{}".format(image_file))
return None
if det and rec:
dt_boxes, rec_res = self.__call__(img)
return [[box.tolist(), res] for box, res in zip(dt_boxes, rec_res)]
elif det and not rec:
dt_boxes, elapse = self.text_detector(img)
if dt_boxes is None:
return None
return [box.tolist() for box in dt_boxes]
else:
if not isinstance(img, list):
img = [img]
rec_res, elapse = self.text_recognizer(img)
return rec_res
requirments.txt
浏览文件 @ 684676f6
......@@ -2,3 +2,5 @@ shapely
imgaug
pyclipper
lmdb
tqdm
numpy
\ No newline at end of file
setup.py 0 → 100644
浏览文件 @ 684676f6
# Copyright (c) 2020 PaddlePaddle Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
from setuptools import setup
from io import open
with open('requirments.txt', encoding="utf-8-sig") as f:
requirements = f.readlines()
requirements.append('tqdm')
def readme():
with open('doc/doc_en/whl.md', encoding="utf-8-sig") as f:
README = f.read()
return README
setup(
name='paddleocr',
packages=['paddleocr'],
package_dir={'paddleocr': ''},
include_package_data=True,
entry_points={"console_scripts": ["paddleocr= paddleocr.paddleocr:main"]},
version='0.0.3',
install_requires=requirements,
license='Apache License 2.0',
description='Awesome OCR toolkits based on PaddlePaddle （8.6M ultra-lightweight pre-trained model, support training and deployment among server, mobile, embeded and IoT devices',
long_description=readme(),
long_description_content_type='text/markdown',
author='Baidu PaddlePaddle',
url='https://github.com/PaddlePaddle/PaddleOCR',
download_url='https://github.com/PaddlePaddle/PaddleOCR.git',
keywords=[
'ocr textdetection textrecognition paddleocr crnn east star-net rosetta ocrlite db chineseocr chinesetextdetection chinesetextrecognition'
],
classifiers=[
'Intended Audience :: Developers', 'Operating System :: OS Independent',
'Natural Language :: Chinese (Simplified)',
'Programming Language :: Python', 'Programming Language :: Python :: 2',
'Programming Language :: Python :: 2.5',
'Programming Language :: Python :: 2.6',
'Programming Language :: Python :: 2.7',
'Programming Language :: Python :: 3',
'Programming Language :: Python :: 3.2',
'Programming Language :: Python :: 3.3',
'Programming Language :: Python :: 3.4',
'Programming Language :: Python :: 3.5',
'Programming Language :: Python :: 3.6',
'Programming Language :: Python :: 3.7', 'Topic :: Utilities'
], )
tools/infer/utility.py
浏览文件 @ 684676f6
......@@ -134,7 +134,12 @@ def resize_img(img, input_size=600):
return im
def draw_ocr(image, boxes, txts, scores, draw_txt=True, drop_score=0.5):
def draw_ocr(image,
boxes,
txts=None,
scores=None,
drop_score=0.5,
font_path="./doc/simfang.ttf"):
"""
Visualize the results of OCR detection and recognition
args:
......@@ -142,23 +147,29 @@ def draw_ocr(image, boxes, txts, scores, draw_txt=True, drop_score=0.5):
boxes(list): boxes with shape(N, 4, 2)
txts(list): the texts
scores(list): txxs corresponding scores
draw_txt(bool): whether draw text or not
drop_score(float): only scores greater than drop_threshold will be visualized
font_path: the path of font which is used to draw text
return(array):
the visualized img
"""
if scores is None:
scores = [1] * len(boxes)
for (box, score) in zip(boxes, scores):
if score < drop_score or math.isnan(score):
box_num = len(boxes)
for i in range(box_num):
if scores is not None and (scores[i] < drop_score or
math.isnan(scores[i])):
continue
box = np.reshape(np.array(box), [-1, 1, 2]).astype(np.int64)
box = np.reshape(np.array(boxes[i]), [-1, 1, 2]).astype(np.int64)
image = cv2.polylines(np.array(image), [box], True, (255, 0, 0), 2)
if draw_txt:
if txts is not None:
img = np.array(resize_img(image, input_size=600))
txt_img = text_visual(
txts, scores, img_h=img.shape[0], img_w=600, threshold=drop_score)
txts,
scores,
img_h=img.shape[0],
img_w=600,
threshold=drop_score,
font_path=font_path)
img = np.concatenate([np.array(img), np.array(txt_img)], axis=1)
return img
return image
......@@ -236,7 +247,12 @@ def str_count(s):
return s_len - math.ceil(en_dg_count / 2)
def text_visual(texts, scores, img_h=400, img_w=600, threshold=0.):
def text_visual(texts,
scores,
img_h=400,
img_w=600,
threshold=0.,
font_path="./doc/simfang.ttf"):
"""
create new blank img and draw txt on it
args:
......@@ -244,6 +260,7 @@ def text_visual(texts, scores, img_h=400, img_w=600, threshold=0.):
scores(list|None): corresponding score of each txt
img_h(int): the height of blank img
img_w(int): the width of blank img
font_path: the path of font which is used to draw text
return(array):
"""
......@@ -262,7 +279,7 @@ def text_visual(texts, scores, img_h=400, img_w=600, threshold=0.):
font_size = 20
txt_color = (0, 0, 0)
font = ImageFont.truetype("./doc/simfang.ttf", font_size, encoding="utf-8")
font = ImageFont.truetype(font_path, font_size, encoding="utf-8")
gap = font_size + 5
txt_img_list = []
......@@ -343,6 +360,6 @@ if __name__ == '__main__':
txts.append(dic['transcription'])
scores.append(round(dic['scores'], 3))
new_img = draw_ocr(image, boxes, txts, scores, draw_txt=True)
new_img = draw_ocr(image, boxes, txts, scores)
cv2.imwrite(img_name, new_img)
Markdown is supported
0% 或 .
You are about to add 0 people to the discussion. Proceed with caution.
先完成此消息的编辑！
想要评论请 注册 或