mirror of
https://github.com/PaddlePaddle/PaddleOCR.git
synced 2026-01-08 04:56:10 +00:00
refine PP-OCRv5_det docs (#15200)
This commit is contained in:
学卿
GitHub
parent
18d0fed174
commit
612b0a6295
@ -124,15 +124,15 @@ for res in output:
|
||||
The output will be:
|
||||
|
||||
```bash
|
||||
{'res': {'input_path': 'general_ocr_001.png', 'page_index': None, 'dt_polys': array([[[ 75, 553],
|
||||
{'res': {'input_path': 'general_ocr_001.png', 'page_index': None, 'dt_polys': array([[[ 75, 549],
|
||||
...,
|
||||
[ 77, 585]],
|
||||
[ 77, 586]],
|
||||
|
||||
...,
|
||||
|
||||
[[ 37, 409],
|
||||
[[ 37, 408],
|
||||
...,
|
||||
[ 39, 453]]], dtype=int16), 'dt_scores': [0.8574396614433001, 0.8343834504056786, 0.8720446452934176, 0.8090656290206919]}}
|
||||
[ 39, 453]]], dtype=int16), 'dt_scores': [0.832930755107492, 0.8186143846140158, 0.8591595100376676, 0.8718863959111733]}}
|
||||
```
|
||||
|
||||
Output parameter meanings:
|
||||
|
||||
@ -125,15 +125,15 @@ for res in output:
|
||||
运行后，得到的结果为：
|
||||
|
||||
```bash
|
||||
{'res': {'input_path': 'general_ocr_001.png', 'page_index': None, 'dt_polys': array([[[ 75, 553],
|
||||
{'res': {'input_path': 'general_ocr_001.png', 'page_index': None, 'dt_polys': array([[[ 75, 549],
|
||||
...,
|
||||
[ 77, 585]],
|
||||
[ 77, 586]],
|
||||
|
||||
...,
|
||||
|
||||
[[ 37, 409],
|
||||
[[ 37, 408],
|
||||
...,
|
||||
[ 39, 453]]], dtype=int16), 'dt_scores': [0.8574396614433001, 0.8343834504056786, 0.8720446452934176, 0.8090656290206919]}}
|
||||
[ 39, 453]]], dtype=int16), 'dt_scores': [0.832930755107492, 0.8186143846140158, 0.8591595100376676, 0.8718863959111733]}}
|
||||
```
|
||||
|
||||
运行结果参数含义如下：
|
||||
|
||||
@ -521,17 +521,16 @@ Before using the general OCR pipeline locally, ensure you have installed the whe
|
||||
Run a single command to quickly test the OCR pipeline:
|
||||
|
||||
```bash
|
||||
# Default: Uses PP-OCRv5 Chinese model
|
||||
paddleocr ocr -i https://paddle-model-ecology.bj.bcebos.com/paddlex/imgs/demo_image/general_ocr_002.png
|
||||
# Default: Uses PP-OCRv5 model
|
||||
paddleocr ocr -i https://paddle-model-ecology.bj.bcebos.com/paddlex/imgs/demo_image/general_ocr_002.png \
|
||||
--use_doc_orientation_classify False \
|
||||
--use_doc_unwarping False \
|
||||
--use_textline_orientation False \
|
||||
--save_path ./output \
|
||||
--device gpu:0
|
||||
|
||||
# Specify PP-OCR version via --ocr_version
|
||||
paddleocr ocr -i ./general_ocr_002.png --ocr_version PP-OCRv4
|
||||
|
||||
# Use GPU for inference via --device
|
||||
paddleocr ocr -i ./general_ocr_002.png --device gpu
|
||||
|
||||
# Disable text line orientation classification via --use_textline_orientation
|
||||
paddleocr ocr -i ./general_ocr_002.png --use_textline_orientation False
|
||||
# Use PP-OCRv4 model by --ocr_version PP-OCRv4
|
||||
paddleocr ocr -i ./general_ocr_002.png --ocr_version PP-OCRv4
|
||||
```
|
||||
|
||||
<details><summary><b>More command-line parameters available. Click to expand for details.</b></summary>
|
||||
@ -819,28 +818,28 @@ paddleocr ocr -i ./general_ocr_002.png --use_textline_orientation False
|
||||
</details>
|
||||
<br />
|
||||
|
||||
Results are printed to the terminal. The default OCR pipeline output is structured as:
|
||||
Results are printed to the terminal:
|
||||
|
||||
```bash
|
||||
{'res': {'input_path': '/root/.paddlex/predict_input/general_ocr_002.png', 'page_index': None, 'model_settings': {'use_doc_preprocessor': True, 'use_textline_orientation': True}, 'doc_preprocessor_res': {'input_path': None, 'page_index': None, 'model_settings': {'use_doc_orientation_classify': True, 'use_doc_unwarping': True}, 'angle': 0}, 'dt_polys': array([[[134, 6],
|
||||
{'res': {'input_path': './general_ocr_002.png', 'page_index': None, 'model_settings': {'use_doc_preprocessor': True, 'use_textline_orientation': False}, 'doc_preprocessor_res': {'input_path': None, 'page_index': None, 'model_settings': {'use_doc_orientation_classify': False, 'use_doc_unwarping': False}, 'angle': -1}, 'dt_polys': array([[[ 3, 10],
|
||||
...,
|
||||
[134, 66]],
|
||||
[ 4, 30]],
|
||||
|
||||
...,
|
||||
|
||||
[[331, 472],
|
||||
[[ 99, 456],
|
||||
...,
|
||||
[331, 494]]], dtype=int16), 'text_det_params': {'limit_side_len': 736, 'limit_type': 'min', 'thresh': 0.3, 'max_side_limit': 4000, 'box_thresh': 0.6, 'unclip_ratio': 1.5}, 'text_type': 'general', 'textline_orientation_angles': array([0, ..., 0]), 'text_rec_score_thresh': 0.0, 'rec_texts': ['登机牌', '', 'BOARDING', 'PASS', '航班', 'FLIGHT', '日期', 'DATE', '舱位', 'CLASS', '序号', 'SERIAL NO.', '座位号', 'SEAT NO', 'MU', '2379', '03DEG', 'M', '035', '12F', '目的地', 'TO', '始发地', 'FROM', '登机口', 'GATE', '登机时间', 'BDT', '福州', 'FUZHOU', 'TAIYUAN', 'G11', '姓名', 'NAME', '身份识别IDNO', 'ZHANGQIWEI', '张祺件', '票号TKTNO', '票价FARE', 'ETKT', '7813699238489/1', '登机口于起飞前10分钟关闭', 'GATES CLOSE 10 MINUTES BEFORE DEPARTURE TIME'], 'rec_scores': array([0.99727374, ..., 0.9353174 ]), 'rec_polys': array([[[134, 6],
|
||||
[ 99, 479]]], dtype=int16), 'text_det_params': {'limit_side_len': 736, 'limit_type': 'min', 'thresh': 0.3, 'max_side_limit': 4000, 'box_thresh': 0.6, 'unclip_ratio': 1.5}, 'text_type': 'general', 'textline_orientation_angles': array([-1, ..., -1]), 'text_rec_score_thresh': 0.0, 'rec_texts': ['www.997700', '', 'Cm', '登机牌', 'BOARDING', 'PASS', 'CLASS', '序号SERIAL NO.', '座位号', 'SEAT NO.', '航班FLIGHT', '日期DATE', '舱位', '', 'W', '035', '12F', 'MU2379', '03DEc', '始发地', 'FROM', '登机口', 'GATE', '登机时间BDT', '目的地TO', '福州', 'TAIYUAN', 'G11', 'FUZHOU', '身份识别IDNO.', '姓名NAME', 'ZHANGQIWEI', '票号TKT NO.', '张祺伟', '票价FARE', 'ETKT7813699238489/1', '登机口于起飞前10分钟关闭 GATESCL0SE10MINUTESBEFOREDEPARTURETIME'], 'rec_scores': array([0.67634439, ..., 0.97416091]), 'rec_polys': array([[[ 3, 10],
|
||||
...,
|
||||
[134, 66]],
|
||||
[ 4, 30]],
|
||||
|
||||
...,
|
||||
|
||||
[[331, 472],
|
||||
[[ 99, 456],
|
||||
...,
|
||||
[331, 494]]], dtype=int16), 'rec_boxes': array([[134, ..., 66],
|
||||
[ 99, 479]]], dtype=int16), 'rec_boxes': array([[ 3, ..., 30],
|
||||
...,
|
||||
[331, ..., 507]], dtype=int16)}}
|
||||
[ 99, ..., 479]], dtype=int16)}}
|
||||
```
|
||||
|
||||
If `save_path` is specified, the visualization results will be saved under `save_path`. The visualization output is shown below:
|
||||
@ -854,11 +853,14 @@ The command-line method is for quick testing. For project integration, you can a
|
||||
```python
|
||||
from paddleocr import PaddleOCR
|
||||
|
||||
ocr = PaddleOCR()
|
||||
# ocr = PaddleOCR(lang="en") # Use English model via `lang`
|
||||
# ocr = PaddleOCR(ocr_version="PP-OCRv4") # Use PP-OCRv4 via `ocr_version`
|
||||
# ocr = PaddleOCR(device="gpu") # Use GPU for inference
|
||||
# ocr = PaddleOCR(use_textline_orientation=False) # Disable text line orientation
|
||||
ocr = PaddleOCR(
|
||||
use_doc_orientation_classify=False, # Disables document orientation classification model via this parameter
|
||||
use_doc_unwarping=False, # Disables text image rectification model via this parameter
|
||||
use_textline_orientation=False, # Disables text line orientation classification model via this parameter
|
||||
)
|
||||
# ocr = PaddleOCR(lang="en") # Uses English model by specifying language parameter
|
||||
# ocr = PaddleOCR(ocr_version="PP-OCRv4") # Uses other PP-OCR versions via version parameter
|
||||
# ocr = PaddleOCR(device="gpu") # Enables GPU acceleration for model inference via device parameter
|
||||
result = ocr.predict("./general_ocr_002.png")
|
||||
for res in result:
|
||||
res.print()
|
||||
|
||||
@ -522,16 +522,15 @@ devanagari_PP-OCRv3_mobile_rec_infer.tar">推理模型</a>/<a href="https://padd
|
||||
|
||||
```bash
|
||||
# 默认使用 PP-OCRv5 模型
|
||||
paddleocr ocr -i https://paddle-model-ecology.bj.bcebos.com/paddlex/imgs/demo_image/general_ocr_002.png
|
||||
paddleocr ocr -i https://paddle-model-ecology.bj.bcebos.com/paddlex/imgs/demo_image/general_ocr_002.png \
|
||||
--use_doc_orientation_classify False \
|
||||
--use_doc_unwarping False \
|
||||
--use_textline_orientation False \
|
||||
--save_path ./output \
|
||||
--device gpu:0
|
||||
|
||||
# 通过 --ocr_version 指定 PP-OCR 其他版本
|
||||
paddleocr ocr -i ./general_ocr_002.png --ocr_version PP-OCRv4
|
||||
|
||||
# 通过 --device 指定模型推理时使用 GPU
|
||||
paddleocr ocr -i ./general_ocr_002.png --device gpu
|
||||
|
||||
# 通过 --use_textline_orientation 指定是否使用文本行方向分类模型
|
||||
paddleocr ocr -i ./general_ocr_002.png --use_textline_orientation False
|
||||
```
|
||||
|
||||
<details><summary><b>命令行支持更多参数设置，点击展开以查看命令行参数的详细说明</b></summary>
|
||||
@ -816,28 +815,28 @@ paddleocr ocr -i ./general_ocr_002.png --use_textline_orientation False
|
||||
</details>
|
||||
<br />
|
||||
|
||||
运行结果会被打印到终端上，默认配置的OCR产线的运行结果如下：
|
||||
运行结果会被打印到终端上：
|
||||
|
||||
```bash
|
||||
{'res': {'input_path': '/root/.paddlex/predict_input/general_ocr_002.png', 'page_index': None, 'model_settings': {'use_doc_preprocessor': True, 'use_textline_orientation': True}, 'doc_preprocessor_res': {'input_path': None, 'page_index': None, 'model_settings': {'use_doc_orientation_classify': True, 'use_doc_unwarping': True}, 'angle': 0}, 'dt_polys': array([[[134, 6],
|
||||
{'res': {'input_path': './general_ocr_002.png', 'page_index': None, 'model_settings': {'use_doc_preprocessor': True, 'use_textline_orientation': False}, 'doc_preprocessor_res': {'input_path': None, 'page_index': None, 'model_settings': {'use_doc_orientation_classify': False, 'use_doc_unwarping': False}, 'angle': -1}, 'dt_polys': array([[[ 3, 10],
|
||||
...,
|
||||
[134, 66]],
|
||||
[ 4, 30]],
|
||||
|
||||
...,
|
||||
|
||||
[[331, 472],
|
||||
[[ 99, 456],
|
||||
...,
|
||||
[331, 494]]], dtype=int16), 'text_det_params': {'limit_side_len': 736, 'limit_type': 'min', 'thresh': 0.3, 'max_side_limit': 4000, 'box_thresh': 0.6, 'unclip_ratio': 1.5}, 'text_type': 'general', 'textline_orientation_angles': array([0, ..., 0]), 'text_rec_score_thresh': 0.0, 'rec_texts': ['登机牌', '', 'BOARDING', 'PASS', '航班', 'FLIGHT', '日期', 'DATE', '舱位', 'CLASS', '序号', 'SERIAL NO.', '座位号', 'SEAT NO', 'MU', '2379', '03DEG', 'M', '035', '12F', '目的地', 'TO', '始发地', 'FROM', '登机口', 'GATE', '登机时间', 'BDT', '福州', 'FUZHOU', 'TAIYUAN', 'G11', '姓名', 'NAME', '身份识别IDNO', 'ZHANGQIWEI', '张祺件', '票号TKTNO', '票价FARE', 'ETKT', '7813699238489/1', '登机口于起飞前10分钟关闭', 'GATES CLOSE 10 MINUTES BEFORE DEPARTURE TIME'], 'rec_scores': array([0.99727374, ..., 0.9353174 ]), 'rec_polys': array([[[134, 6],
|
||||
[ 99, 479]]], dtype=int16), 'text_det_params': {'limit_side_len': 736, 'limit_type': 'min', 'thresh': 0.3, 'max_side_limit': 4000, 'box_thresh': 0.6, 'unclip_ratio': 1.5}, 'text_type': 'general', 'textline_orientation_angles': array([-1, ..., -1]), 'text_rec_score_thresh': 0.0, 'rec_texts': ['www.997700', '', 'Cm', '登机牌', 'BOARDING', 'PASS', 'CLASS', '序号SERIAL NO.', '座位号', 'SEAT NO.', '航班FLIGHT', '日期DATE', '舱位', '', 'W', '035', '12F', 'MU2379', '03DEc', '始发地', 'FROM', '登机口', 'GATE', '登机时间BDT', '目的地TO', '福州', 'TAIYUAN', 'G11', 'FUZHOU', '身份识别IDNO.', '姓名NAME', 'ZHANGQIWEI', '票号TKT NO.', '张祺伟', '票价FARE', 'ETKT7813699238489/1', '登机口于起飞前10分钟关闭 GATESCL0SE10MINUTESBEFOREDEPARTURETIME'], 'rec_scores': array([0.67634439, ..., 0.97416091]), 'rec_polys': array([[[ 3, 10],
|
||||
...,
|
||||
[134, 66]],
|
||||
[ 4, 30]],
|
||||
|
||||
...,
|
||||
|
||||
[[331, 472],
|
||||
[[ 99, 456],
|
||||
...,
|
||||
[331, 494]]], dtype=int16), 'rec_boxes': array([[134, ..., 66],
|
||||
[ 99, 479]]], dtype=int16), 'rec_boxes': array([[ 3, ..., 30],
|
||||
...,
|
||||
[331, ..., 507]], dtype=int16)}}
|
||||
[ 99, ..., 479]], dtype=int16)}}
|
||||
```
|
||||
|
||||
若指定了`save_path`，则会保存可视化结果在`save_path`下。可视化结果如下：
|
||||
@ -851,11 +850,14 @@ paddleocr ocr -i ./general_ocr_002.png --use_textline_orientation False
|
||||
```python
|
||||
from paddleocr import PaddleOCR
|
||||
|
||||
ocr = PaddleOCR()
|
||||
ocr = PaddleOCR(
|
||||
use_doc_orientation_classify=False, # 通过 use_doc_orientation_classify 参数指定不使用文档方向分类模型
|
||||
use_doc_unwarping=False, # 通过 use_doc_unwarping 参数指定不使用文本图像矫正模型
|
||||
use_textline_orientation=False, # 通过 use_textline_orientation 参数指定不使用文本行方向分类模型
|
||||
)
|
||||
# ocr = PaddleOCR(lang="en") # 通过 lang 参数来使用英文模型
|
||||
# ocr = PaddleOCR(ocr_version="PP-OCRv4") # 通过 ocr_version 参数来使用 PP-OCR 其他版本
|
||||
# ocr = PaddleOCR(device="gpu") # 通过 device 参数使得在模型推理时使用 GPU
|
||||
# ocr = PaddleOCR(use_textline_orientation=False) # 通过 device 参数指定不使用文本行方向分类模型
|
||||
result = ocr.predict("./general_ocr_002.png")
|
||||
for res in result:
|
||||
res.print()
|
||||
@ -1607,7 +1609,6 @@ for i, res in enumerate(result["ocrResults"]):
|
||||
|
||||
通用 OCR 产线包含若干模块，模型产线的效果如果不及预期，可能来自于其中任何一个模块。您可以对识别效果差的图片进行分析，进而确定是哪个模块存在问题，并参考以下表格中对应的微调教程链接进行模型微调。
|
||||
|
||||
|
||||
<table>
|
||||
<thead>
|
||||
<tr>
|
||||
|
||||
Loading…
x
Reference in New Issue
Block a user