PaddleX/latest/module_usage/tutorials/ocr_modules/textline_orientation_classification

[giscus[bot]]

PaddleX/latest/module_usage/tutorials/ocr_modules/textline_orientation_classification

https://paddlepaddle.github.io/PaddleX/latest/module_usage/tutorials/ocr_modules/textline_orientation_classification.html

[sbuques]

弧形文字可以吗？

[cuicheng01]

比如是什么样的弧形文字呢？

[sbuques]

[cuicheng01]

这种图像原则上是不可以的

[AndyJado]

paddleOCR的切片操作能在单模块里调用吗？

[chenqan]

有办法识别 90度 或 270度的文本， 涉及到版面区域检测模块，没有文本没有矫正，版面区域检测的效果不太好

[nainai123]

我也想问支持多分类吗

[liuhongen1234567]

您好，这个文本行分类的目的就是为了将 90度或270度的文本旋转带文本可以检测的水平方向。多分类可以参考文档图像多标签分类模块文档 https://paddlepaddle.github.io/PaddleX/latest/module_usage/tutorials/cv_modules/image_multilabel_classification.html

[ooooevan]

这个文档里面用到的 main.py 内容是什么？

[liuhongen1234567]

您好，main.py 是PaddleX中的主文件，需要下载并安装PaddleX，才能使用

[13691126988]

请问为什么呢？
(paddleocr) chenrh@node63:~/PaddleX-3.0.2$ python main.py -c paddlex/configs/modules/textline_orientation/PP-LCNet_x1_0_textline_ori.yaml \

-o Global.mode=train \
-o Global.dataset_dir=./dataset/textline_orientation_example_data

Traceback (most recent call last):
File "/home/chenrh/PaddleX-3.0.2/paddlex/repo_apis/base/config.py", line 30, in _create_config
model_info = get_registered_model_info(model_name)
File "/home/chenrh/PaddleX-3.0.2/paddlex/repo_apis/base/register.py", line 78, in query
return self._table[prim_key]
KeyError: 'PP-LCNet_x1_0_textline_ori'

The above exception was the direct cause of the following exception:

Traceback (most recent call last):
File "/home/chenrh/PaddleX-3.0.2/paddlex/utils/result_saver.py", line 28, in wrap
result = func(self, *args, **kwargs)
File "/home/chenrh/PaddleX-3.0.2/paddlex/engine.py", line 41, in run
self._model.train()
File "/home/chenrh/PaddleX-3.0.2/paddlex/model.py", line 118, in train
trainer = build_trainer(self._config)
File "/home/chenrh/PaddleX-3.0.2/paddlex/modules/base/trainer.py", line 44, in build_trainer
return BaseTrainer.get(model_name)(config)
File "/home/chenrh/PaddleX-3.0.2/paddlex/modules/base/trainer.py", line 66, in init
self.pdx_config, self.pdx_model = build_model(
File "/home/chenrh/PaddleX-3.0.2/paddlex/modules/base/build_model.py", line 31, in build_model
config = Config(model_name, config_path)
File "/home/chenrh/PaddleX-3.0.2/paddlex/repo_apis/base/config.py", line 32, in _create_config
raise UnsupportedParamError(
paddlex.utils.errors.others.UnsupportedParamError: 'PP-LCNet_x1_0_textline_ori' is not a registered model name.
(paddleocr) chenrh@node63:/PaddleX-3.0.2$ paddleocr --version
paddleocr 3.0.2
(paddleocr) chenrh@node63:/PaddleX-3.0.2$ pip show paddlex
Name: paddlex
Version: 3.0.2
Summary: Low-code development tool based on PaddlePaddle.
Home-page:
Author: PaddlePaddle Authors
Author-email:
License: Apache 2.0
Location: /home/chenrh/PaddleX-3.0.2
Editable project location: /home/chenrh/PaddleX-3.0.2
Requires: chardet, colorlog, filelock, huggingface_hub, numpy, packaging, pandas, pillow, prettytable, py-cpuinfo, pydantic, PyYAML, requests, ruamel.yaml, typing-extensions, ujson
Required-by: paddleocr

[13691126988]

尝试了一下3.0.1和3.0.2，都不可以进行训练，都提示paddlex.utils.errors.others.UnsupportedParamError: 'PP-LCNet_x1_0_textline_ori' is not a registered model name.

[yyyyyxie]

我也遇到了这个问题，train和evaluate都不可以，不过mode=predict倒是可以

[yyyyyxie]

刚刚又试了一下，是环境问题，现在可以了

[liuhongen1234567]

您好，训练需要安装对应插件，比如： PP-LCNet_x1_0_textline_ori 需要通过下面的命令安装分类插件：

paddlex --install PaddleClas

[gtfaiwxm]

进行韩文的文本行方向分类时，效果不佳，想问一下主要支持哪几种语言的文本行方向分类？