mirror of
https://github.com/PaddlePaddle/PaddleOCR.git
synced 2025-06-03 21:53:39 +08:00
Add deployment docs and enhance CLI (#15117)
* Add serving and hpi docs * Optimize CLI logging info * Update interface * Add on-device deployment and onnx model conversion docs * Enhance CLI * _gen->_iter * Fix CLI help message * Update table_recognition_v2 and PP-StructureV3 interfaces * Update installation doc * Update interface * Update interface * Add logging doc * Update default values --------- Co-authored-by: cuicheng01 <45199522+cuicheng01@users.noreply.github.com>
This commit is contained in:
Lin Manhui
GitHub
parent
3d594881e4
commit
b25dcaae0e
91
docs/deployment/high_performance_inference.md
Normal file
91
docs/deployment/high_performance_inference.md
Normal file
@ -0,0 +1,91 @@
|
||||
# 高性能推理
|
||||
|
||||
在实际生产环境中，许多应用对部署策略的性能指标（尤其是响应速度）有着较严苛的标准，以确保系统的高效运行与用户体验的流畅性。PaddleOCR 提供高性能推理能力，让用户无需关注复杂的配置和底层细节，一键提升模型的推理速度。具体而言，PaddleOCR 的高性能推理功能能够：
|
||||
|
||||
- 结合先验知识自动选择合适的推理后端（Paddle Inference、OpenVINO、ONNX Runtime、TensorRT等），并配置加速策略（如增大推理线程数、设置 FP16 精度推理）；
|
||||
- 根据需要自动将飞桨静态图模型转换为 ONNX 格式，以使用更优的推理后端实现加速。
|
||||
|
||||
本文档主要介绍高性能推理功能的安装与使用方法。
|
||||
|
||||
## 1. 前置条件
|
||||
|
||||
## 1.1 安装高性能推理依赖
|
||||
|
||||
通过 PaddleOCR CLI 安装高性能推理所需依赖：
|
||||
|
||||
```bash
|
||||
paddleocr install_hpi_deps {设备类型}
|
||||
```
|
||||
|
||||
支持的设备类型包括：
|
||||
|
||||
- `cpu`：仅使用 CPU 推理。目前支持 Linux 系统、x86-64 架构处理器、Python 3.8-3.12。
|
||||
- `gpu`：使用 CPU 或 NVIDIA GPU 推理。目前支持 Linux 系统、x86-64 架构处理器、Python 3.8-3.12。请查看下一小节的详细说明。
|
||||
|
||||
同一环境中只应该存在一种设备类型的依赖。对于 Windows 系统，目前建议在 Docker 容器或者 [WSL](https://learn.microsoft.com/zh-cn/windows/wsl/install) 环境中安装。
|
||||
|
||||
**推荐使用飞桨官方 Docker 镜像安装高性能推理依赖。** 各设备类型对应的镜像如下：
|
||||
|
||||
- `cpu`：`ccr-2vdh3abv-pub.cnc.bj.baidubce.com/paddlepaddle/paddle:3.0.0`
|
||||
- `gpu`：
|
||||
- CUDA 11.8：`ccr-2vdh3abv-pub.cnc.bj.baidubce.com/paddlepaddle/paddle:3.0.0-gpu-cuda11.8-cudnn8.9-trt8.6`
|
||||
|
||||
## 1.2 GPU 环境详细说明
|
||||
|
||||
首先，需要确保环境中安装有符合要求的 CUDA 与 cuDNN。目前 PaddleOCR 仅支持与 CUDA 11.8 + cuDNN 8.9 兼容的 CUDA 和 cuDNN版本。以下分别是 CUDA 11.8 和 cuDNN 8.9 的安装说明文档：
|
||||
|
||||
- [安装 CUDA 11.8](https://developer.nvidia.com/cuda-11-8-0-download-archive)
|
||||
- [安装 cuDNN 8.9](https://docs.nvidia.com/deeplearning/cudnn/archives/cudnn-890/install-guide/index.html)
|
||||
|
||||
如果使用飞桨官方镜像，则镜像中的 CUDA 和 cuDNN 版本已经是满足要求的，无需额外安装。
|
||||
|
||||
如果通过 pip 安装飞桨，通常 CUDA、cuDNN 的相关 Python 包将被自动安装。在这种情况下，**仍需要通过安装非 Python 专用的 CUDA 与 cuDNN**。同时，建议安装的 CUDA 和 cuDNN 版本与环境中存在的 Python 包版本保持一致，以避免不同版本的库共存导致的潜在问题。可以通过如下方式可以查看 CUDA 和 cuDNN 相关 Python 包的版本：
|
||||
|
||||
```bash
|
||||
# CUDA 相关 Python 包版本
|
||||
pip list | grep nvidia-cuda
|
||||
# cuDNN 相关 Python 包版本
|
||||
pip list | grep nvidia-cudnn
|
||||
```
|
||||
|
||||
其次，需确保环境中安装有符合要求的 TensorRT。目前 PaddleOCR 仅支持 TensorRT 8.6.1.6。如果使用飞桨官方镜像，可执行如下命令安装 TensorRT wheel 包：
|
||||
|
||||
```bash
|
||||
python -m pip install /usr/local/TensorRT-*/python/tensorrt-*-cp310-none-linux_x86_64.whl
|
||||
```
|
||||
|
||||
对于其他环境，请参考 [TensorRT 文档](https://docs.nvidia.com/deeplearning/tensorrt/archives/index.html) 安装 TensorRT。示例如下：
|
||||
|
||||
```bash
|
||||
# 下载 TensorRT tar 文件
|
||||
wget https://developer.nvidia.com/downloads/compute/machine-learning/tensorrt/secure/8.6.1/tars/TensorRT-8.6.1.6.Linux.x86_64-gnu.cuda-11.8.tar.gz
|
||||
# 解压 TensorRT tar 文件
|
||||
tar xvf TensorRT-8.6.1.6.Linux.x86_64-gnu.cuda-11.8.tar.gz
|
||||
# 安装 TensorRT wheel 包
|
||||
python -m pip install TensorRT-8.6.1.6/python/tensorrt-8.6.1-cp310-none-linux_x86_64.whl
|
||||
# 添加 TensorRT 的 `lib` 目录的绝对路径到 LD_LIBRARY_PATH 中
|
||||
export LD_LIBRARY_PATH="$LD_LIBRARY_PATH:TensorRT-8.6.1.6/lib"
|
||||
```
|
||||
|
||||
## 2. 执行高性能推理
|
||||
|
||||
对于 PaddleOCR CLI，指定 `--enable_hpi` 为 `True` 即可执行高性能推理。例如：
|
||||
|
||||
```bash
|
||||
paddleocr ocr --enable_hpi True ...
|
||||
```
|
||||
|
||||
对于 PaddleOCR Python API，在初始化产线对象或者模块对象时，设置 `enable_hpi` 为 `True` 即可在调用推理方法时执行高性能推理。例如：
|
||||
|
||||
```python
|
||||
from paddleocr import PaddleOCR
|
||||
pipeline = PaddleOCR(enable_hpi=True)
|
||||
result = pipeline.predict(...)
|
||||
```
|
||||
|
||||
## 3. 说明
|
||||
|
||||
1. 对于部分模型，在首次执行高性能推理时，可能需要花费较长时间完成推理引擎的构建。推理引擎相关信息将在第一次构建完成后被缓存在模型目录，后续可复用缓存中的内容以提升初始化速度。
|
||||
2. 目前，由于使用的不是静态图格式模型、存在不支持算子等原因，部分模型可能无法获得推理加速。
|
||||
3. 在进行高性能推理时，PaddleOCR 会自动处理模型格式的转换，并尽可能选择最优的推理后端。同时，PaddleOCR 也支持用户指定 ONNX 模型。有关如何飞桨静态图模型转换为 ONNX 格式，可参考 [获取 ONNX 模型](./obtaining_onnx_models.md)。
|
||||
4. PaddleOCR 的高性能推理能力依托于 PaddleX 及其高性能推理插件。通过传入自定义 PaddleX 产线配置文件，可以对推理后端等进行配置。请参考 [使用 PaddleX 产线配置文件](../paddleocr_and_paddlex.md#3-使用-paddlex-产线配置文件) 和 [PaddleX 高性能推理指南](https://paddlepaddle.github.io/PaddleX/3.0/pipeline_deploy/high_performance_inference.html#22) 了解如何调整高性能推理配置。
|
||||
48
docs/deployment/obtaining_onnx_models.md
Normal file
48
docs/deployment/obtaining_onnx_models.md
Normal file
@ -0,0 +1,48 @@
|
||||
# 获取 ONNX 模型
|
||||
|
||||
PaddleOCR 提供了丰富的预训练模型，这些模型均采用飞桨的静态图格式进行存储。若需在部署阶段使用 ONNX 格式的模型，可借助 PaddleX 提供的 Paddle2ONNX 插件进行转换。关于 PaddleX 及其与 PaddleOCR 之间的关系，请参考 [PaddleOCR 与 PaddleX 的区别与联系](../paddleocr_and_paddlex.md#1-paddleocr-与-paddlex-的区别与联系)。
|
||||
|
||||
首先，执行如下命令，通过 PaddleX CLI 安装 PaddleX 的 Paddle2ONNX 插件：
|
||||
|
||||
```bash
|
||||
paddlex --install paddle2onnx
|
||||
```
|
||||
|
||||
然后，执行如下命令完成模型转换：
|
||||
|
||||
```bash
|
||||
paddlex \
|
||||
--paddle2onnx \ # 使用paddle2onnx功能
|
||||
--paddle_model_dir /your/paddle_model/dir \ # 指定 Paddle 模型所在的目录
|
||||
--onnx_model_dir /your/onnx_model/output/dir \ # 指定转换后 ONNX 模型的输出目录
|
||||
--opset_version 7 # 指定要使用的 ONNX opset 版本
|
||||
```
|
||||
|
||||
参数说明如下：
|
||||
|
||||
<table>
|
||||
<thead>
|
||||
<tr>
|
||||
<th>参数</th>
|
||||
<th>类型</th>
|
||||
<th>描述</th>
|
||||
</tr>
|
||||
</thead>
|
||||
<tbody>
|
||||
<tr>
|
||||
<td>paddle_model_dir</td>
|
||||
<td>str</td>
|
||||
<td>包含 Paddle 模型的目录。</td>
|
||||
</tr>
|
||||
<tr>
|
||||
<td>onnx_model_dir</td>
|
||||
<td>str</td>
|
||||
<td>ONNX 模型的输出目录，可以与 Paddle 模型目录相同。默认为 <code>onnx</code>。</td>
|
||||
</tr>
|
||||
<tr>
|
||||
<td>opset_version</td>
|
||||
<td>int</td>
|
||||
<td>使用的 ONNX opset 版本。当使用低版本 opset 无法完成转换时，将自动选择更高版本的 opset 进行转换。默认为 <code>7</code>。</td>
|
||||
</tr>
|
||||
</tbody>
|
||||
</table>
|
||||
3
docs/deployment/on_device_deployment.md
Normal file
3
docs/deployment/on_device_deployment.md
Normal file
@ -0,0 +1,3 @@
|
||||
# 端侧部署
|
||||
|
||||
PaddleOCR 模型可通过 [PaddleX 端侧部署方案](https://paddlepaddle.github.io/PaddleX/3.0/pipeline_deploy/edge_deploy.html) 实现端侧部署。关于 PaddleX 及其与 PaddleOCR 之间的关系，请参考 [PaddleOCR 与 PaddleX 的区别与联系](../paddleocr_and_paddlex.md#1-paddleocr-与-paddlex-的区别与联系)。
|
||||
93
docs/deployment/serving.md
Normal file
93
docs/deployment/serving.md
Normal file
@ -0,0 +1,93 @@
|
||||
# 服务化部署
|
||||
|
||||
服务化部署是实际生产环境中常见的一种部署形式。通过将推理功能封装为服务，客户端可以通过网络请求来访问这些服务，以获取推理结果。PaddleOCR 推荐用户使用 [PaddleX](https://github.com/PaddlePaddle/PaddleX) 进行服务化部署。请阅读 [PaddleOCR 与 PaddleX 的区别与联系](../paddleocr_and_paddlex.md#1-paddleocr-与-paddlex-的区别与联系) 了解 PaddleOCR 与 PaddleX 的关系。
|
||||
|
||||
PaddleX 提供以下服务化部署方案：
|
||||
|
||||
- **基础服务化部署**：简单易用的服务化部署方案，开发成本低。
|
||||
- **高稳定性服务化部署**：基于 [NVIDIA Triton Inference Server](https://developer.nvidia.com/triton-inference-server) 打造。与基础服务化部署相比，该方案提供更高的稳定性，并允许用户调整配置以优化性能。
|
||||
|
||||
**建议首先使用基础服务化部署方案进行快速验证**，然后根据实际需要，评估是否尝试更复杂的方案。
|
||||
|
||||
## 1. 基础服务化部署
|
||||
|
||||
### 1.1 安装依赖
|
||||
|
||||
执行如下命令，通过 PaddleX CLI 安装 PaddleX 服务化部署插件：
|
||||
|
||||
```bash
|
||||
paddlex --install serving
|
||||
```
|
||||
|
||||
### 1.2 运行服务器
|
||||
|
||||
通过 PaddleX CLI 运行服务器：
|
||||
|
||||
```bash
|
||||
paddlex --serve --pipeline {PaddleX 产线注册名或产线配置文件路径} [{其他命令行选项}]
|
||||
```
|
||||
|
||||
以通用 OCR 产线为例：
|
||||
|
||||
```bash
|
||||
paddlex --serve --pipeline OCR
|
||||
```
|
||||
|
||||
可以看到类似以下展示的信息：
|
||||
|
||||
```text
|
||||
INFO: Started server process [63108]
|
||||
INFO: Waiting for application startup.
|
||||
INFO: Application startup complete.
|
||||
INFO: Uvicorn running on http://0.0.0.0:8080 (Press CTRL+C to quit)
|
||||
```
|
||||
|
||||
如需调整配置（如模型路径、batch size、部署设备等），可指定 `--pipeline` 为自定义配置文件。请参考 [PaddleOCR 与 PaddleX](../advanced/paddleocr_and_paddlex.md) 了解 PaddleOCR 产线与 PaddleX 产线注册名的对应关系，以及 PaddleX 产线配置文件的获取与修改方式。
|
||||
|
||||
与服务化部署相关的命令行选项如下：
|
||||
|
||||
<table>
|
||||
<thead>
|
||||
<tr>
|
||||
<th>名称</th>
|
||||
<th>说明</th>
|
||||
</tr>
|
||||
</thead>
|
||||
<tbody>
|
||||
<tr>
|
||||
<td><code>--pipeline</code></td>
|
||||
<td>PaddleX 产线注册名或产线配置文件路径。</td>
|
||||
</tr>
|
||||
<tr>
|
||||
<td><code>--device</code></td>
|
||||
<td>产线部署设备。默认为 <code>cpu</code>（如 GPU 不可用）或 <code>gpu</code>（如 GPU 可用）。</td>
|
||||
</tr>
|
||||
<tr>
|
||||
<td><code>--host</code></td>
|
||||
<td>服务器绑定的主机名或 IP 地址。默认为 <code>0.0.0.0</code>。</td>
|
||||
</tr>
|
||||
<tr>
|
||||
<td><code>--port</code></td>
|
||||
<td>服务器监听的端口号。默认为 <code>8080</code>。</td>
|
||||
</tr>
|
||||
<tr>
|
||||
<td><code>--use_hpip</code></td>
|
||||
<td>如果指定，则使用高性能推理。</td>
|
||||
</tr>
|
||||
<tr>
|
||||
<td><code>--hpi_config</code></td>
|
||||
<td>高性能推理配置。请参考 <a href="https://paddlepaddle.github.io/PaddleX/3.0/pipeline_deploy/high_performance_inference.html#22">PaddleX 高性能推理指南</a> 了解更多信息。</td>
|
||||
</tr>
|
||||
</tbody>
|
||||
</table>
|
||||
</table>
|
||||
|
||||
### 1.3 调用服务
|
||||
|
||||
PaddleX 产线使用教程中的 <b>“开发集成/部署”</b> 部分提供了服务的 API 参考与多语言调用示例。在 [PaddleX模型产线使用概览](https://paddlepaddle.github.io/PaddleX/3.0/pipeline_usage/pipeline_develop_guide.html) 中可以找到各产线的使用教程。
|
||||
|
||||
## 2. 高稳定性服务化部署
|
||||
|
||||
请参考 [PaddleX 服务化部署指南](https://paddlepaddle.github.io/PaddleX/3.0/pipeline_deploy/serving.html#2)。在 [使用 PaddleX 产线配置文件](../paddleocr_and_paddlex.md#3-使用-paddlex-产线配置文件) 中，可以了解关于 PaddleX 产线配置文件的更多信息。
|
||||
|
||||
需要说明的是，由于缺乏细粒度优化等原因，当前 PaddleOCR 提供的高稳定性服务化部署方案在性能上可能不及 2.x 版本基于 PaddleServing 的方案；但该新方案已对飞桨 3.0 框架提供了全面支持，我们也将持续优化，后续考虑推出性能更优的部署方案。
|
||||
22
docs/logging.md
Normal file
22
docs/logging.md
Normal file
@ -0,0 +1,22 @@
|
||||
# 日志
|
||||
|
||||
本文档主要介绍如何配置 PaddleOCR 推理包的日志系统。需要注意的是，PaddleOCR 推理包与训练脚本使用的是不同的日志系统，本文档不涉及训练脚本所使用的日志系统的配置方法。
|
||||
|
||||
PaddleOCR 构建了一个基于 Python [`logging` 标准库](https://docs.python.org/zh-cn/3/library/logging.html#module-logging) 的集中式日志系统。换言之，PaddleOCR 使用唯一的日志记录器（logger），可通过 `paddleocr.logger` 访问和配置。
|
||||
|
||||
默认情况下，PaddleOCR 的日志级别设为 `ERROR`，这意味着仅当日志级别为 `ERROR` 或更高（如 `CRITICAL`）时，日志信息才会输出。PaddleOCR 同时为该日志记录器配置了一个 `StreamHandler`，将日志输出到标准错误流，并将记录器的 `propagate` 属性设为 `False`，以避免日志信息传递到其父记录器。
|
||||
|
||||
若希望禁止 PaddleOCR 对日志系统的自动配置行为，可将环境变量 `DISABLE_AUTO_LOGGING_CONFIG` 设为 `1`。此时，PaddleOCR 将不会对日志记录器进行任何额外配置。
|
||||
|
||||
如需更灵活地定制日志行为，可参考 `logging` 标准库的相关文档。以下是一个将日志写入文件的示例：
|
||||
|
||||
```python
|
||||
import logging
|
||||
from paddleocr import logger
|
||||
|
||||
# 将日志写入文件 `paddleocr.log`
|
||||
fh = logging.FileHandler("paddleocr.log")
|
||||
logger.addHandler(fh)
|
||||
```
|
||||
|
||||
请注意，PaddleOCR 依赖的其他库（如 [PaddleX](./paddleocr_and_paddlex.md)）拥有各自独立的日志系统，以上配置不会影响这些库的日志输出。
|
||||
69
docs/paddleocr_and_paddlex.md
Normal file
69
docs/paddleocr_and_paddlex.md
Normal file
@ -0,0 +1,69 @@
|
||||
# PaddleOCR 与 PaddleX
|
||||
|
||||
[PaddleX](https://github.com/PaddlePaddle/PaddleX) 是一款基于飞桨框架构建的低代码开发工具，集成了众多开箱即用的预训练模型，支持模型从训练到推理的全流程开发，兼容多款国内外主流硬件，助力 AI 开发者在产业实践中高效落地。
|
||||
|
||||
PaddleOCR 在推理部署方面基于 PaddleX 构建，二者在该环节可实现无缝协同。在安装 PaddleOCR 时，PaddleX 也将作为其依赖一并安装。此外，PaddleOCR 与 PaddleX 在产线名称等方面也保持一致。对于快速体验，如果只使用基础配置，用户通常无需了解 PaddleX 的具体概念；但在涉及高级配置、服务化部署等使用场景时，了解 PaddleX 的相关知识将有所帮助。
|
||||
|
||||
本文档将介绍 PaddleOCR 与 PaddleX 之间的关系，并说明如何协同使用这两个工具。
|
||||
|
||||
## 1. PaddleOCR 与 PaddleX 的区别与联系
|
||||
|
||||
PaddleOCR 与 PaddleX 在定位和功能上各有侧重：PaddleOCR 专注于 OCR 相关任务，而 PaddleX 则覆盖了包括时序预测、人脸识别等在内的多种任务类型。此外，PaddleX 提供了丰富的基础设施，具备多模型组合推理的底层能力，能够以统一且灵活的方式接入不同模型，支持构建复杂的模型产线。
|
||||
|
||||
PaddleOCR 在推理部署环节充分复用了 PaddleX 的能力，具体包括：
|
||||
|
||||
- PaddleOCR 在模型推理、前后处理及多模型组合等底层能力上，主要依赖于 PaddleX。
|
||||
- PaddleOCR 的高性能推理能力通过 PaddleX 的 Paddle2ONNX 插件及高性能推理插件实现。
|
||||
- PaddleOCR 的服务化部署方案基于 PaddleX 的实现。
|
||||
|
||||
## 2. PaddleOCR 产线与 PaddleX 产线注册名的对应关系
|
||||
|
||||
| PaddleOCR 产线 | PaddleX 产线注册名 |
|
||||
| --- | --- |
|
||||
| 通用 OCR | `OCR` |
|
||||
| 通用版面解析 v3 | `PP-StructureV3` |
|
||||
| 文档场景信息抽取 v4 | `PP-ChatOCRv4-doc` |
|
||||
| 通用表格识别 v2 | `table_recognition_v2` |
|
||||
| 公式识别 | `formula_recognition` |
|
||||
| 印章文本识别 | `seal_recognition` |
|
||||
| 文档图像预处理 | `doc_preprocessor` |
|
||||
| 文档理解 | `doc_understanding` |
|
||||
|
||||
## 3. 使用 PaddleX 产线配置文件
|
||||
|
||||
在推理部署阶段，PaddleOCR 支持导出和加载 PaddleX 的产线配置文件。用户可通过编辑配置文件，对推理部署相关参数进行深度配置。
|
||||
|
||||
### 3.1 导出产线配置文件
|
||||
|
||||
可调用 PaddleOCR 产线对象的 `export_paddlex_config_to_yaml` 方法，将当前产线配置导出为 YAML 文件。示例如下：
|
||||
|
||||
```python
|
||||
from paddleocr import PaddleOCR
|
||||
|
||||
pipeline = PaddleOCR()
|
||||
pipeline.export_paddlex_config_to_yaml("ocr_config.yaml")
|
||||
```
|
||||
|
||||
上述代码会在工作目录下生成名为 `ocr_config.yaml` 的产线配置文件。
|
||||
|
||||
### 3.2 编辑产线配置文件
|
||||
|
||||
导出的 PaddleX 产线配置文件不仅包含 PaddleOCR CLI 和 Python API 支持的参数，还可进行更多高级配置。请在 [PaddleX模型产线使用概览](https://paddlepaddle.github.io/PaddleX/3.0/pipeline_usage/pipeline_develop_guide.html) 中找到对应的产线使用教程，参考其中的详细说明，根据需求调整各项配置。
|
||||
|
||||
### 3.3 在 CLI 中加载产线配置文件
|
||||
|
||||
通过 `--paddlex_config` 参数指定 PaddleX 产线配置文件的路径，PaddleOCR 会读取其中的内容作为产线的默认配置。示例如下：
|
||||
|
||||
```bash
|
||||
paddleocr ocr --paddlex_config ocr_config.yaml ...
|
||||
```
|
||||
|
||||
### 3.4 在 Python API 中加载产线配置文件
|
||||
|
||||
初始化产线对象时，可通过 `paddlex_config` 参数传入 PaddleX 产线配置文件路径或配置字典，PaddleOCR 会将其作为默认配置。示例如下：
|
||||
|
||||
```python
|
||||
from paddleocr import PaddleOCR
|
||||
|
||||
pipeline = PaddleOCR(paddlex_config="ocr_config.yaml")
|
||||
```
|
||||
@ -1,85 +0,0 @@
|
||||
---
|
||||
comments: true
|
||||
---
|
||||
|
||||
## Quick Installation
|
||||
|
||||
After testing, PaddleOCR can run on glibc 2.23. You can also test other glibc versions or install glibc 2.23 for the best compatibility.
|
||||
|
||||
PaddleOCR working environment:
|
||||
|
||||
- PaddlePaddle > 2.0.0
|
||||
- Python 3
|
||||
- glibc 2.23
|
||||
|
||||
It is recommended to use the docker provided by us to run PaddleOCR. Please refer to the docker tutorial [link](https://www.runoob.com/docker/docker-tutorial.html/).
|
||||
|
||||
*If you want to directly run the prediction code on Mac or Windows, you can start from step 2.*
|
||||
|
||||
### 1. (Recommended) Prepare a docker environment
|
||||
|
||||
For the first time you use this docker image, it will be downloaded automatically. Please be patient.
|
||||
|
||||
```bash linenums="1"
|
||||
# Switch to the working directory
|
||||
cd /home/Projects
|
||||
# You need to create a docker container for the first run, and do not need to run the current command when you run it again
|
||||
# Create a docker container named ppocr and map the current directory to the /paddle directory of the container
|
||||
|
||||
#If using CPU, use docker instead of nvidia-docker to create docker
|
||||
sudo docker run --name ppocr -v $PWD:/paddle --network=host -it paddlepaddle/paddle:latest-dev-cuda10.1-cudnn7-gcc82 /bin/bash
|
||||
```
|
||||
|
||||
With CUDA10, please run the following command to create a container.
|
||||
It is recommended to set a shared memory greater than or equal to 32G through the --shm-size parameter:
|
||||
|
||||
```bash linenums="1"
|
||||
sudo nvidia-docker run --name ppocr -v $PWD:/paddle --shm-size=64G --network=host -it paddlepaddle/paddle:latest-dev-cuda10.1-cudnn7-gcc82 /bin/bash
|
||||
```
|
||||
|
||||
You can also visit [DockerHub](https://hub.docker.com/r/paddlepaddle/paddle/tags/) to get the image that fits your machine.
|
||||
|
||||
```bash linenums="1"
|
||||
# ctrl+P+Q to exit docker, to re-enter docker using the following command:
|
||||
sudo docker container exec -it ppocr /bin/bash
|
||||
```
|
||||
|
||||
### 2. Install PaddlePaddle 2.0
|
||||
|
||||
```bash linenums="1"
|
||||
pip3 install --upgrade pip
|
||||
|
||||
# If you have cuda9 or cuda10 installed on your machine, please run the following command to install
|
||||
python3 -m pip install paddlepaddle-gpu==2.0.0 -i https://mirror.baidu.com/pypi/simple
|
||||
|
||||
# If you only have cpu on your machine, please run the following command to install
|
||||
python3 -m pip install paddlepaddle==2.0.0 -i https://mirror.baidu.com/pypi/simple
|
||||
```
|
||||
|
||||
For more software version requirements, please refer to the instructions in [Installation Document](https://www.paddlepaddle.org.cn/install/quick) for operation.
|
||||
|
||||
### 3. Clone PaddleOCR repo
|
||||
|
||||
```bash linenums="1"
|
||||
# Recommend
|
||||
git clone https://github.com/PaddlePaddle/PaddleOCR
|
||||
|
||||
# If you cannot pull successfully due to network problems, you can switch to the mirror hosted on Gitee:
|
||||
|
||||
git clone https://gitee.com/paddlepaddle/PaddleOCR
|
||||
|
||||
# Note: The mirror on Gitee may not keep in synchronization with the latest update with the project on GitHub. There might be a delay of 3-5 days. Please try GitHub at first.
|
||||
```
|
||||
|
||||
### 4. Install third-party libraries
|
||||
|
||||
```bash linenums="1"
|
||||
cd PaddleOCR
|
||||
pip3 install -r requirements.txt
|
||||
```
|
||||
|
||||
If you getting this error `OSError: [WinError 126] The specified module could not be found` when you install shapely on windows.
|
||||
|
||||
Please try to download Shapely whl file from [http://www.lfd.uci.edu/~gohlke/pythonlibs/#shapely](http://www.lfd.uci.edu/~gohlke/pythonlibs/#shapely).
|
||||
|
||||
Reference: [Solve shapely installation on windows](https://stackoverflow.com/questions/44398265/install-shapely-oserror-winerror-126-the-specified-module-could-not-be-found)
|
||||
@ -2,75 +2,49 @@
|
||||
comments: true
|
||||
---
|
||||
|
||||
## 快速安装
|
||||
# 安装
|
||||
|
||||
经测试PaddleOCR可在glibc 2.23上运行，您也可以测试其他glibc版本或安装glic 2.23
|
||||
PaddleOCR 工作环境
|
||||
# 1. 安装飞桨框架
|
||||
|
||||
- PaddlePaddle 2.0.0
|
||||
- python3
|
||||
- glibc 2.23
|
||||
- cuDNN 7.6+ (GPU)
|
||||
请参考 [飞桨官网](https://www.paddlepaddle.org.cn/install/quick?docurl=/documentation/docs/zh/develop/install/pip/linux-pip.html) 安装 `3.0` 及以上版本的飞桨框架。**推荐使用飞桨官方 Docker 镜像。**
|
||||
|
||||
建议使用我们提供的docker运行PaddleOCR，有关docker、nvidia-docker使用请参考[链接](https://www.runoob.com/docker/docker-tutorial.html/)。
|
||||
# 2. 安装 PaddleOCR
|
||||
|
||||
*如您希望使用 mac 或 windows直接运行预测代码，可以从第2步开始执行。*
|
||||
如果只希望使用 PaddleOCR 的推理功能，请参考 [安装推理包](#21-安装推理包)；如果希望进行模型训练、导出等，请参考 [安装训练依赖](#22-安装训练依赖)。在同一环境中安装推理包和训练依赖是允许的，无需进行环境隔离。
|
||||
|
||||
### 1. （建议）准备docker环境
|
||||
## 2.1 安装推理包
|
||||
|
||||
第一次使用这个镜像，会自动下载该镜像，请耐心等待
|
||||
从 PyPI 安装最新版本 PaddleOCR 推理包：
|
||||
|
||||
```bash linenums="1"
|
||||
# 切换到工作目录下
|
||||
cd /home/Projects
|
||||
# 首次运行需创建一个docker容器，再次运行时不需要运行当前命令
|
||||
# 创建一个名字为ppocr的docker容器，并将当前目录映射到容器的/paddle目录下
|
||||
|
||||
如果您希望在CPU环境下使用docker，使用docker而不是nvidia-docker创建docker
|
||||
sudo docker run --name ppocr -v $PWD:/paddle --network=host -it paddlepaddle/paddle:latest-dev-cuda10.1-cudnn7-gcc82 /bin/bash
|
||||
|
||||
如果使用CUDA10，请运行以下命令创建容器，设置docker容器共享内存shm-size为64G，建议设置32G以上
|
||||
sudo nvidia-docker run --name ppocr -v $PWD:/paddle --shm-size=64G --network=host -it paddlepaddle/paddle:latest-dev-cuda10.1-cudnn7-gcc82 /bin/bash
|
||||
|
||||
您也可以访问[DockerHub](https://hub.docker.com/r/paddlepaddle/paddle/tags/)获取与您机器适配的镜像。
|
||||
|
||||
# ctrl+P+Q可退出docker 容器，重新进入docker 容器使用如下命令
|
||||
sudo docker container exec -it ppocr /bin/bash
|
||||
```bash
|
||||
python -m pip install paddleocr
|
||||
```
|
||||
|
||||
### 2. 安装PaddlePaddle 2.0
|
||||
或者从源码安装（默认为开发分支）：
|
||||
|
||||
```bash linenums="1"
|
||||
pip3 install --upgrade pip
|
||||
|
||||
# 如果您的机器安装的是CUDA9或CUDA10，请运行以下命令安装
|
||||
python3 -m pip install paddlepaddle-gpu==2.0.0 -i https://mirror.baidu.com/pypi/simple
|
||||
|
||||
# 如果您的机器是CPU，请运行以下命令安装
|
||||
python3 -m pip install paddlepaddle==2.0.0 -i https://mirror.baidu.com/pypi/simple
|
||||
|
||||
# 更多的版本需求，请参照[安装文档](https://www.paddlepaddle.org.cn/install/quick)中的说明进行操作。
|
||||
```bash
|
||||
python -m pip install "git+https://github.com/PaddlePaddle/PaddleOCR.git"
|
||||
```
|
||||
|
||||
### 3. 克隆PaddleOCR repo代码
|
||||
## 2.2 安装训练依赖
|
||||
|
||||
```bash linenums="1"
|
||||
#【推荐】
|
||||
要进行模型训练、导出等，需要首先将仓库克隆到本地：
|
||||
|
||||
```bash
|
||||
# 推荐方式
|
||||
git clone https://github.com/PaddlePaddle/PaddleOCR
|
||||
|
||||
# 如果因为网络问题无法pull成功，也可选择使用码云上的托管：
|
||||
# （可选）切换到指定分支
|
||||
git checkout release/3.0
|
||||
|
||||
# 如果因为网络问题无法克隆成功，也可选择使用码云上的仓库：
|
||||
git clone https://gitee.com/paddlepaddle/PaddleOCR
|
||||
|
||||
# 注：码云托管代码可能无法实时同步本github项目更新，存在3~5天延时，请优先使用推荐方式。
|
||||
# 注：码云托管代码可能无法实时同步本 GitHub 项目更新，存在3~5天延时，请优先使用推荐方式。
|
||||
```
|
||||
|
||||
### 4. 安装第三方库
|
||||
执行如下命令安装依赖：
|
||||
|
||||
```bash linenums="1"
|
||||
cd PaddleOCR
|
||||
pip3 install -r requirements.txt
|
||||
```bash
|
||||
python -m pip install -r requirements.txt
|
||||
```
|
||||
|
||||
注意，windows环境下，建议从[这里](https://www.lfd.uci.edu/~gohlke/pythonlibs/#shapely)下载shapely安装包完成安装，
|
||||
直接通过pip安装的shapely库可能出现`[winRrror 126] 找不到指定模块的问题`。
|
||||
|
||||
@ -36,6 +36,7 @@ from ._pipelines import (
|
||||
SealRecognition,
|
||||
TableRecognitionPipelineV2,
|
||||
)
|
||||
from ._utils.logging import logger
|
||||
from ._version import version as __version__
|
||||
|
||||
__all__ = [
|
||||
@ -59,5 +60,6 @@ __all__ = [
|
||||
"PPStructureV3",
|
||||
"SealRecognition",
|
||||
"TableRecognitionPipelineV2",
|
||||
"logger",
|
||||
"__version__",
|
||||
]
|
||||
|
||||
@ -13,8 +13,10 @@
|
||||
# limitations under the License.
|
||||
|
||||
import argparse
|
||||
import logging
|
||||
import subprocess
|
||||
import sys
|
||||
|
||||
import warnings
|
||||
|
||||
from ._models import (
|
||||
@ -42,7 +44,8 @@ from ._pipelines import (
|
||||
TableRecognitionPipelineV2,
|
||||
)
|
||||
from ._version import version
|
||||
from .utils.deprecation import CLIDeprecationWarning
|
||||
from ._utils.deprecation import CLIDeprecationWarning
|
||||
from ._utils.logging import logger
|
||||
|
||||
|
||||
def _register_pipelines(subparsers):
|
||||
@ -110,6 +113,7 @@ def _execute(args):
|
||||
|
||||
|
||||
def main():
|
||||
logger.setLevel(logging.INFO)
|
||||
warnings.filterwarnings("default", category=CLIDeprecationWarning)
|
||||
args = _parse_args()
|
||||
_execute(args)
|
||||
|
||||
@ -24,7 +24,7 @@ from ._constants import (
|
||||
DEFAULT_USE_TENSORRT,
|
||||
SUPPORTED_PRECISION_LIST,
|
||||
)
|
||||
from .utils.cli import str2bool
|
||||
from ._utils.cli import str2bool
|
||||
|
||||
|
||||
def parse_common_args(kwargs, *, default_enable_hpi):
|
||||
|
||||
@ -14,7 +14,7 @@
|
||||
|
||||
import abc
|
||||
|
||||
from ..utils.cli import (
|
||||
from .._utils.cli import (
|
||||
add_simple_inference_args,
|
||||
get_subcommand_args,
|
||||
perform_simple_inference,
|
||||
|
||||
@ -14,7 +14,7 @@
|
||||
|
||||
import abc
|
||||
|
||||
from ..utils.cli import (
|
||||
from .._utils.cli import (
|
||||
add_simple_inference_args,
|
||||
get_subcommand_args,
|
||||
perform_simple_inference,
|
||||
|
||||
@ -23,7 +23,7 @@ from .._common_args import (
|
||||
prepare_common_init_args,
|
||||
)
|
||||
from .._mkldnn_blocklists import MODEL_MKLDNN_BLOCKLIST
|
||||
from ..utils.logging import logger
|
||||
from .._utils.logging import logger
|
||||
|
||||
_DEFAULT_ENABLE_HPI = False
|
||||
|
||||
@ -59,10 +59,11 @@ class PaddleXPredictorWrapper(metaclass=abc.ABCMeta):
|
||||
def default_model_name(self):
|
||||
raise NotImplementedError
|
||||
|
||||
def predict_iter(self, *args, **kwargs):
|
||||
return self.paddlex_predictor.predict(*args, **kwargs)
|
||||
|
||||
def predict(self, *args, **kwargs):
|
||||
result = []
|
||||
for res in self.paddlex_predictor.predict(*args, **kwargs):
|
||||
result.append(res)
|
||||
result = list(self.predict_iter(*args, **kwargs))
|
||||
return result
|
||||
|
||||
@classmethod
|
||||
|
||||
@ -12,7 +12,9 @@
|
||||
# See the License for the specific language governing permissions and
|
||||
# limitations under the License.
|
||||
|
||||
from ..utils.cli import (
|
||||
from paddlex.utils.pipeline_arguments import custom_type
|
||||
|
||||
from .._utils.cli import (
|
||||
add_simple_inference_args,
|
||||
get_subcommand_args,
|
||||
perform_simple_inference,
|
||||
@ -50,7 +52,10 @@ class DocVLMSubcommandExecutor(PredictorCLISubcommandExecutor):
|
||||
return "doc_vlm"
|
||||
|
||||
def _update_subparser(self, subparser):
|
||||
add_simple_inference_args(subparser)
|
||||
add_simple_inference_args(
|
||||
subparser,
|
||||
input_help='Input dict, e.g. `{"image": "https://paddle-model-ecology.bj.bcebos.com/paddlex/imgs/demo_image/medal_table.png", "query": "Recognize this table"}`.',
|
||||
)
|
||||
|
||||
def execute_with_args(self, args):
|
||||
params = get_subcommand_args(args)
|
||||
|
||||
@ -12,7 +12,7 @@
|
||||
# See the License for the specific language governing permissions and
|
||||
# limitations under the License.
|
||||
|
||||
from ..utils.cli import (
|
||||
from .._utils.cli import (
|
||||
add_simple_inference_args,
|
||||
get_subcommand_args,
|
||||
perform_simple_inference,
|
||||
|
||||
@ -12,7 +12,7 @@
|
||||
# See the License for the specific language governing permissions and
|
||||
# limitations under the License.
|
||||
|
||||
from ..utils.cli import (
|
||||
from .._utils.cli import (
|
||||
add_simple_inference_args,
|
||||
get_subcommand_args,
|
||||
perform_simple_inference,
|
||||
|
||||
@ -12,7 +12,7 @@
|
||||
# See the License for the specific language governing permissions and
|
||||
# limitations under the License.
|
||||
|
||||
from ..utils.cli import (
|
||||
from .._utils.cli import (
|
||||
add_simple_inference_args,
|
||||
get_subcommand_args,
|
||||
perform_simple_inference,
|
||||
|
||||
@ -12,7 +12,7 @@
|
||||
# See the License for the specific language governing permissions and
|
||||
# limitations under the License.
|
||||
|
||||
from ..utils.cli import (
|
||||
from .._utils.cli import (
|
||||
add_simple_inference_args,
|
||||
get_subcommand_args,
|
||||
perform_simple_inference,
|
||||
|
||||
@ -12,7 +12,7 @@
|
||||
# See the License for the specific language governing permissions and
|
||||
# limitations under the License.
|
||||
|
||||
from ..utils.cli import (
|
||||
from .._utils.cli import (
|
||||
add_simple_inference_args,
|
||||
get_subcommand_args,
|
||||
perform_simple_inference,
|
||||
|
||||
@ -17,6 +17,7 @@ import abc
|
||||
import yaml
|
||||
from paddlex import create_pipeline
|
||||
from paddlex.inference import load_pipeline_config
|
||||
from paddlex.utils.config import AttrDict
|
||||
|
||||
from .._abstract import CLISubcommandExecutor
|
||||
from .._common_args import (
|
||||
@ -25,7 +26,7 @@ from .._common_args import (
|
||||
prepare_common_init_args,
|
||||
)
|
||||
from .._mkldnn_blocklists import PIPELINE_MKLDNN_BLOCKLIST
|
||||
from ..utils.logging import logger
|
||||
from .._utils.logging import logger
|
||||
|
||||
_DEFAULT_ENABLE_HPI = None
|
||||
|
||||
@ -40,6 +41,14 @@ def _merge_dicts(d1, d2):
|
||||
return res
|
||||
|
||||
|
||||
def _to_plain_dict(d):
|
||||
res = d.copy()
|
||||
for k, v in d.items():
|
||||
if isinstance(v, AttrDict):
|
||||
res[k] = _to_plain_dict(v)
|
||||
return res
|
||||
|
||||
|
||||
class PaddleXPipelineWrapper(metaclass=abc.ABCMeta):
|
||||
def __init__(
|
||||
self,
|
||||
@ -70,7 +79,8 @@ class PaddleXPipelineWrapper(metaclass=abc.ABCMeta):
|
||||
|
||||
def export_paddlex_config_to_yaml(self, yaml_path):
|
||||
with open(yaml_path, "w", encoding="utf-8") as f:
|
||||
yaml.safe_dump(self._merged_paddlex_config, f)
|
||||
config = _to_plain_dict(self._merged_paddlex_config)
|
||||
yaml.safe_dump(config, f)
|
||||
|
||||
@classmethod
|
||||
@abc.abstractmethod
|
||||
@ -83,7 +93,7 @@ class PaddleXPipelineWrapper(metaclass=abc.ABCMeta):
|
||||
def _get_merged_paddlex_config(self):
|
||||
if self._paddlex_config is None:
|
||||
config = load_pipeline_config(self._paddlex_pipeline_name)
|
||||
elif isinstance(self._config, str):
|
||||
elif isinstance(self._paddlex_config, str):
|
||||
config = load_pipeline_config(self._paddlex_config)
|
||||
else:
|
||||
config = self._paddlex_config
|
||||
|
||||
@ -12,7 +12,7 @@
|
||||
# See the License for the specific language governing permissions and
|
||||
# limitations under the License.
|
||||
|
||||
from ..utils.cli import (
|
||||
from .._utils.cli import (
|
||||
add_simple_inference_args,
|
||||
get_subcommand_args,
|
||||
perform_simple_inference,
|
||||
@ -48,6 +48,19 @@ class DocPreprocessor(PaddleXPipelineWrapper):
|
||||
def _paddlex_pipeline_name(self):
|
||||
return "doc_preprocessor"
|
||||
|
||||
def predict_iter(
|
||||
self,
|
||||
input,
|
||||
*,
|
||||
use_doc_orientation_classify=None,
|
||||
use_doc_unwarping=None,
|
||||
):
|
||||
return self.paddlex_pipeline.predict(
|
||||
input,
|
||||
use_doc_orientation_classify=use_doc_orientation_classify,
|
||||
use_doc_unwarping=use_doc_unwarping,
|
||||
)
|
||||
|
||||
def predict(
|
||||
self,
|
||||
input,
|
||||
@ -55,14 +68,13 @@ class DocPreprocessor(PaddleXPipelineWrapper):
|
||||
use_doc_orientation_classify=None,
|
||||
use_doc_unwarping=None,
|
||||
):
|
||||
result = []
|
||||
for res in self.paddlex_pipeline.predict(
|
||||
input,
|
||||
use_doc_orientation_classify=use_doc_orientation_classify,
|
||||
use_doc_unwarping=use_doc_unwarping,
|
||||
):
|
||||
result.append(res)
|
||||
return result
|
||||
return list(
|
||||
self.predict_iter(
|
||||
input,
|
||||
use_doc_orientation_classify=use_doc_orientation_classify,
|
||||
use_doc_unwarping=use_doc_unwarping,
|
||||
)
|
||||
)
|
||||
|
||||
@classmethod
|
||||
def get_cli_subcommand_executor(cls):
|
||||
|
||||
@ -12,15 +12,15 @@
|
||||
# See the License for the specific language governing permissions and
|
||||
# limitations under the License.
|
||||
|
||||
from ..utils.cli import (
|
||||
from paddlex.utils.pipeline_arguments import custom_type
|
||||
|
||||
from .._utils.cli import (
|
||||
add_simple_inference_args,
|
||||
get_subcommand_args,
|
||||
perform_simple_inference,
|
||||
str2bool,
|
||||
)
|
||||
from .base import PaddleXPipelineWrapper, PipelineCLISubcommandExecutor
|
||||
from .utils import create_config_from_structure
|
||||
from paddlex.utils.pipeline_arguments import custom_type
|
||||
|
||||
|
||||
class DocUnderstanding(PaddleXPipelineWrapper):
|
||||
@ -43,15 +43,15 @@ class DocUnderstanding(PaddleXPipelineWrapper):
|
||||
def _paddlex_pipeline_name(self):
|
||||
return "doc_understanding"
|
||||
|
||||
def predict_iter(self, input, **kwargs):
|
||||
return self.paddlex_pipeline.predict(input, **kwargs)
|
||||
|
||||
def predict(
|
||||
self,
|
||||
input,
|
||||
**kwargs,
|
||||
):
|
||||
result = []
|
||||
for res in self.paddlex_pipeline.predict(input, **kwargs):
|
||||
result.append(res)
|
||||
return result
|
||||
return list(self.predict_iter(input, **kwargs))
|
||||
|
||||
@classmethod
|
||||
def get_cli_subcommand_executor(cls):
|
||||
@ -73,7 +73,6 @@ class DocUnderstanding(PaddleXPipelineWrapper):
|
||||
|
||||
|
||||
class DocUnderstandingCLISubcommandExecutor(PipelineCLISubcommandExecutor):
|
||||
|
||||
input_validator = staticmethod(custom_type(dict))
|
||||
|
||||
@property
|
||||
@ -81,7 +80,10 @@ class DocUnderstandingCLISubcommandExecutor(PipelineCLISubcommandExecutor):
|
||||
return "doc_understanding"
|
||||
|
||||
def _update_subparser(self, subparser):
|
||||
add_simple_inference_args(subparser)
|
||||
add_simple_inference_args(
|
||||
subparser,
|
||||
input_help='Input dict, e.g. `{"image": "https://paddle-model-ecology.bj.bcebos.com/paddlex/imgs/demo_image/medal_table.png", "query": "Recognize this table"}`.',
|
||||
)
|
||||
|
||||
subparser.add_argument(
|
||||
"--doc_understanding_model_name",
|
||||
|
||||
@ -12,13 +12,12 @@
|
||||
# See the License for the specific language governing permissions and
|
||||
# limitations under the License.
|
||||
|
||||
from ..utils.cli import (
|
||||
from .._utils.cli import (
|
||||
add_simple_inference_args,
|
||||
get_subcommand_args,
|
||||
perform_simple_inference,
|
||||
str2bool,
|
||||
)
|
||||
from ..utils.logging import logger
|
||||
from .base import PaddleXPipelineWrapper, PipelineCLISubcommandExecutor
|
||||
from .utils import create_config_from_structure
|
||||
|
||||
@ -58,6 +57,33 @@ class FormulaRecognitionPipeline(PaddleXPipelineWrapper):
|
||||
def _paddlex_pipeline_name(self):
|
||||
return "formula_recognition"
|
||||
|
||||
def predict_iter(
|
||||
self,
|
||||
input,
|
||||
*,
|
||||
use_layout_detection=None,
|
||||
use_doc_orientation_classify=None,
|
||||
use_doc_unwarping=None,
|
||||
layout_det_res=None,
|
||||
layout_threshold=None,
|
||||
layout_nms=None,
|
||||
layout_unclip_ratio=None,
|
||||
layout_merge_bboxes_mode=None,
|
||||
**kwargs,
|
||||
):
|
||||
return self.paddlex_pipeline.predict(
|
||||
input,
|
||||
use_layout_detection=use_layout_detection,
|
||||
use_doc_orientation_classify=use_doc_orientation_classify,
|
||||
use_doc_unwarping=use_doc_unwarping,
|
||||
layout_det_res=layout_det_res,
|
||||
layout_threshold=layout_threshold,
|
||||
layout_nms=layout_nms,
|
||||
layout_unclip_ratio=layout_unclip_ratio,
|
||||
layout_merge_bboxes_mode=layout_merge_bboxes_mode,
|
||||
**kwargs,
|
||||
)
|
||||
|
||||
def predict(
|
||||
self,
|
||||
input,
|
||||
@ -72,21 +98,20 @@ class FormulaRecognitionPipeline(PaddleXPipelineWrapper):
|
||||
layout_merge_bboxes_mode=None,
|
||||
**kwargs,
|
||||
):
|
||||
result = []
|
||||
for res in self.paddlex_pipeline.predict(
|
||||
input,
|
||||
use_layout_detection=use_layout_detection,
|
||||
use_doc_orientation_classify=use_doc_orientation_classify,
|
||||
use_doc_unwarping=use_doc_unwarping,
|
||||
layout_det_res=layout_det_res,
|
||||
layout_threshold=layout_threshold,
|
||||
layout_nms=layout_nms,
|
||||
layout_unclip_ratio=layout_unclip_ratio,
|
||||
layout_merge_bboxes_mode=layout_merge_bboxes_mode,
|
||||
**kwargs,
|
||||
):
|
||||
result.append(res)
|
||||
return result
|
||||
return list(
|
||||
self.predict_iter(
|
||||
input,
|
||||
use_layout_detection=use_layout_detection,
|
||||
use_doc_orientation_classify=use_doc_orientation_classify,
|
||||
use_doc_unwarping=use_doc_unwarping,
|
||||
layout_det_res=layout_det_res,
|
||||
layout_threshold=layout_threshold,
|
||||
layout_nms=layout_nms,
|
||||
layout_unclip_ratio=layout_unclip_ratio,
|
||||
layout_merge_bboxes_mode=layout_merge_bboxes_mode,
|
||||
**kwargs,
|
||||
)
|
||||
)
|
||||
|
||||
@classmethod
|
||||
def get_cli_subcommand_executor(cls):
|
||||
|
||||
@ -18,18 +18,18 @@
|
||||
|
||||
import sys
|
||||
|
||||
from ..utils.cli import (
|
||||
from .._utils.cli import (
|
||||
add_simple_inference_args,
|
||||
get_subcommand_args,
|
||||
perform_simple_inference,
|
||||
str2bool,
|
||||
)
|
||||
from ..utils.deprecation import (
|
||||
from .._utils.deprecation import (
|
||||
DeprecatedOptionAction,
|
||||
deprecated,
|
||||
warn_deprecated_param,
|
||||
)
|
||||
from ..utils.logging import logger
|
||||
from .._utils.logging import logger
|
||||
from .base import PaddleXPipelineWrapper, PipelineCLISubcommandExecutor
|
||||
from .utils import create_config_from_structure
|
||||
|
||||
@ -147,6 +147,33 @@ class PaddleOCR(PaddleXPipelineWrapper):
|
||||
def _paddlex_pipeline_name(self):
|
||||
return "OCR"
|
||||
|
||||
def predict_iter(
|
||||
self,
|
||||
input,
|
||||
*,
|
||||
use_doc_orientation_classify=None,
|
||||
use_doc_unwarping=None,
|
||||
use_textline_orientation=None,
|
||||
text_det_limit_side_len=None,
|
||||
text_det_limit_type=None,
|
||||
text_det_thresh=None,
|
||||
text_det_box_thresh=None,
|
||||
text_det_unclip_ratio=None,
|
||||
text_rec_score_thresh=None,
|
||||
):
|
||||
return self.paddlex_pipeline.predict(
|
||||
input,
|
||||
use_doc_orientation_classify=use_doc_orientation_classify,
|
||||
use_doc_unwarping=use_doc_unwarping,
|
||||
use_textline_orientation=use_textline_orientation,
|
||||
text_det_limit_side_len=text_det_limit_side_len,
|
||||
text_det_limit_type=text_det_limit_type,
|
||||
text_det_thresh=text_det_thresh,
|
||||
text_det_box_thresh=text_det_box_thresh,
|
||||
text_det_unclip_ratio=text_det_unclip_ratio,
|
||||
text_rec_score_thresh=text_rec_score_thresh,
|
||||
)
|
||||
|
||||
def predict(
|
||||
self,
|
||||
input,
|
||||
@ -161,21 +188,20 @@ class PaddleOCR(PaddleXPipelineWrapper):
|
||||
text_det_unclip_ratio=None,
|
||||
text_rec_score_thresh=None,
|
||||
):
|
||||
result = []
|
||||
for res in self.paddlex_pipeline.predict(
|
||||
input,
|
||||
use_doc_orientation_classify=use_doc_orientation_classify,
|
||||
use_doc_unwarping=use_doc_unwarping,
|
||||
use_textline_orientation=use_textline_orientation,
|
||||
text_det_limit_side_len=text_det_limit_side_len,
|
||||
text_det_limit_type=text_det_limit_type,
|
||||
text_det_thresh=text_det_thresh,
|
||||
text_det_box_thresh=text_det_box_thresh,
|
||||
text_det_unclip_ratio=text_det_unclip_ratio,
|
||||
text_rec_score_thresh=text_rec_score_thresh,
|
||||
):
|
||||
result.append(res)
|
||||
return result
|
||||
return list(
|
||||
self.predict_iter(
|
||||
input,
|
||||
use_doc_orientation_classify=use_doc_orientation_classify,
|
||||
use_doc_unwarping=use_doc_unwarping,
|
||||
use_textline_orientation=use_textline_orientation,
|
||||
text_det_limit_side_len=text_det_limit_side_len,
|
||||
text_det_limit_type=text_det_limit_type,
|
||||
text_det_thresh=text_det_thresh,
|
||||
text_det_box_thresh=text_det_box_thresh,
|
||||
text_det_unclip_ratio=text_det_unclip_ratio,
|
||||
text_rec_score_thresh=text_rec_score_thresh,
|
||||
)
|
||||
)
|
||||
|
||||
@deprecated("Please use `predict` instead.")
|
||||
def ocr(self, img, **kwargs):
|
||||
|
||||
@ -12,7 +12,7 @@
|
||||
# See the License for the specific language governing permissions and
|
||||
# limitations under the License.
|
||||
|
||||
from ..utils.cli import (
|
||||
from .._utils.cli import (
|
||||
get_subcommand_args,
|
||||
str2bool,
|
||||
)
|
||||
@ -43,7 +43,6 @@ class PPChatOCRv4Doc(PaddleXPipelineWrapper):
|
||||
seal_text_recognition_batch_size=None,
|
||||
use_doc_orientation_classify=None,
|
||||
use_doc_unwarping=None,
|
||||
use_general_ocr=None,
|
||||
use_seal_recognition=None,
|
||||
use_table_recognition=None,
|
||||
layout_threshold=None,
|
||||
@ -78,13 +77,12 @@ class PPChatOCRv4Doc(PaddleXPipelineWrapper):
|
||||
def _paddlex_pipeline_name(self):
|
||||
return "PP-ChatOCRv4-doc"
|
||||
|
||||
def visual_predict(
|
||||
def visual_predict_iter(
|
||||
self,
|
||||
input,
|
||||
*,
|
||||
use_doc_orientation_classify=None,
|
||||
use_doc_unwarping=None,
|
||||
use_general_ocr=None,
|
||||
use_seal_recognition=None,
|
||||
use_table_recognition=None,
|
||||
layout_threshold=None,
|
||||
@ -105,12 +103,10 @@ class PPChatOCRv4Doc(PaddleXPipelineWrapper):
|
||||
seal_rec_score_thresh=None,
|
||||
**kwargs,
|
||||
):
|
||||
result = []
|
||||
for res in self.paddlex_pipeline.visual_predict(
|
||||
return self.paddlex_pipeline.visual_predict(
|
||||
input,
|
||||
use_doc_orientation_classify=use_doc_orientation_classify,
|
||||
use_doc_unwarping=use_doc_unwarping,
|
||||
use_general_ocr=use_general_ocr,
|
||||
use_seal_recognition=use_seal_recognition,
|
||||
use_table_recognition=use_table_recognition,
|
||||
layout_threshold=layout_threshold,
|
||||
@ -130,9 +126,60 @@ class PPChatOCRv4Doc(PaddleXPipelineWrapper):
|
||||
seal_det_unclip_ratio=seal_det_unclip_ratio,
|
||||
seal_rec_score_thresh=seal_rec_score_thresh,
|
||||
**kwargs,
|
||||
):
|
||||
result.append(res)
|
||||
return result
|
||||
)
|
||||
|
||||
def visual_predict(
|
||||
self,
|
||||
input,
|
||||
*,
|
||||
use_doc_orientation_classify=None,
|
||||
use_doc_unwarping=None,
|
||||
use_seal_recognition=None,
|
||||
use_table_recognition=None,
|
||||
layout_threshold=None,
|
||||
layout_nms=None,
|
||||
layout_unclip_ratio=None,
|
||||
layout_merge_bboxes_mode=None,
|
||||
text_det_limit_side_len=None,
|
||||
text_det_limit_type=None,
|
||||
text_det_thresh=None,
|
||||
text_det_box_thresh=None,
|
||||
text_det_unclip_ratio=None,
|
||||
text_rec_score_thresh=None,
|
||||
seal_det_limit_side_len=None,
|
||||
seal_det_limit_type=None,
|
||||
seal_det_thresh=None,
|
||||
seal_det_box_thresh=None,
|
||||
seal_det_unclip_ratio=None,
|
||||
seal_rec_score_thresh=None,
|
||||
**kwargs,
|
||||
):
|
||||
return list(
|
||||
self.visual_predict_iter(
|
||||
input,
|
||||
use_doc_orientation_classify=use_doc_orientation_classify,
|
||||
use_doc_unwarping=use_doc_unwarping,
|
||||
use_seal_recognition=use_seal_recognition,
|
||||
use_table_recognition=use_table_recognition,
|
||||
layout_threshold=layout_threshold,
|
||||
layout_nms=layout_nms,
|
||||
layout_unclip_ratio=layout_unclip_ratio,
|
||||
layout_merge_bboxes_mode=layout_merge_bboxes_mode,
|
||||
text_det_limit_side_len=text_det_limit_side_len,
|
||||
text_det_limit_type=text_det_limit_type,
|
||||
text_det_thresh=text_det_thresh,
|
||||
text_det_box_thresh=text_det_box_thresh,
|
||||
text_det_unclip_ratio=text_det_unclip_ratio,
|
||||
text_rec_score_thresh=text_rec_score_thresh,
|
||||
seal_det_limit_side_len=seal_det_limit_side_len,
|
||||
seal_det_limit_type=seal_det_limit_type,
|
||||
seal_det_thresh=seal_det_thresh,
|
||||
seal_det_box_thresh=seal_det_box_thresh,
|
||||
seal_det_unclip_ratio=seal_det_unclip_ratio,
|
||||
seal_rec_score_thresh=seal_rec_score_thresh,
|
||||
**kwargs,
|
||||
)
|
||||
)
|
||||
|
||||
def build_vector(
|
||||
self,
|
||||
@ -269,9 +316,6 @@ class PPChatOCRv4Doc(PaddleXPipelineWrapper):
|
||||
"SubPipelines.LayoutParser.SubPipelines.DocPreprocessor.use_doc_unwarping": self._params[
|
||||
"use_doc_unwarping"
|
||||
],
|
||||
"SubPipelines.LayoutParser.use_general_ocr": self._params[
|
||||
"use_general_ocr"
|
||||
],
|
||||
"SubPipelines.LayoutParser.use_seal_recognition": self._params[
|
||||
"use_seal_recognition"
|
||||
],
|
||||
@ -468,11 +512,6 @@ class PPChatOCRv4DocCLISubcommandExecutor(PipelineCLISubcommandExecutor):
|
||||
type=str2bool,
|
||||
help="Whether to use the text image unwarping model.",
|
||||
)
|
||||
subparser.add_argument(
|
||||
"--use_general_ocr",
|
||||
type=str2bool,
|
||||
help="Whether to use general OCR.",
|
||||
)
|
||||
subparser.add_argument(
|
||||
"--use_seal_recognition",
|
||||
type=str2bool,
|
||||
|
||||
@ -12,7 +12,7 @@
|
||||
# See the License for the specific language governing permissions and
|
||||
# limitations under the License.
|
||||
|
||||
from ..utils.cli import (
|
||||
from .._utils.cli import (
|
||||
add_simple_inference_args,
|
||||
get_subcommand_args,
|
||||
perform_simple_inference,
|
||||
@ -31,6 +31,11 @@ class PPStructureV3(PaddleXPipelineWrapper):
|
||||
layout_nms=None,
|
||||
layout_unclip_ratio=None,
|
||||
layout_merge_bboxes_mode=None,
|
||||
chart_recognition_model_name=None,
|
||||
chart_recognition_model_dir=None,
|
||||
chart_recognition_batch_size=None,
|
||||
region_detection_model_name=None,
|
||||
region_detection_model_dir=None,
|
||||
doc_orientation_classify_model_name=None,
|
||||
doc_orientation_classify_model_dir=None,
|
||||
doc_unwarping_model_name=None,
|
||||
@ -75,10 +80,11 @@ class PPStructureV3(PaddleXPipelineWrapper):
|
||||
formula_recognition_batch_size=None,
|
||||
use_doc_orientation_classify=None,
|
||||
use_doc_unwarping=None,
|
||||
use_general_ocr=None,
|
||||
use_seal_recognition=None,
|
||||
use_table_recognition=None,
|
||||
use_formula_recognition=None,
|
||||
use_chart_recognition=None,
|
||||
use_region_detection=None,
|
||||
**kwargs,
|
||||
):
|
||||
params = locals().copy()
|
||||
@ -92,16 +98,17 @@ class PPStructureV3(PaddleXPipelineWrapper):
|
||||
def _paddlex_pipeline_name(self):
|
||||
return "PP-StructureV3"
|
||||
|
||||
def predict(
|
||||
def predict_iter(
|
||||
self,
|
||||
input,
|
||||
use_doc_orientation_classify=None,
|
||||
use_doc_unwarping=None,
|
||||
use_doc_orientation_classify=False,
|
||||
use_doc_unwarping=False,
|
||||
use_textline_orientation=None,
|
||||
use_general_ocr=None,
|
||||
use_seal_recognition=None,
|
||||
use_table_recognition=None,
|
||||
use_formula_recognition=None,
|
||||
use_chart_recognition=False,
|
||||
use_region_detection=None,
|
||||
layout_threshold=None,
|
||||
layout_nms=None,
|
||||
layout_unclip_ratio=None,
|
||||
@ -118,21 +125,24 @@ class PPStructureV3(PaddleXPipelineWrapper):
|
||||
seal_det_box_thresh=None,
|
||||
seal_det_unclip_ratio=None,
|
||||
seal_rec_score_thresh=None,
|
||||
use_table_cells_ocr_results=None,
|
||||
use_e2e_wired_table_rec_model=None,
|
||||
use_e2e_wireless_table_rec_model=None,
|
||||
use_wired_table_cells_trans_to_html=False,
|
||||
use_wireless_table_cells_trans_to_html=False,
|
||||
use_table_orientation_classify=True,
|
||||
use_ocr_results_with_table_cells=True,
|
||||
use_e2e_wired_table_rec_model=False,
|
||||
use_e2e_wireless_table_rec_model=True,
|
||||
**kwargs,
|
||||
):
|
||||
result = []
|
||||
for res in self.paddlex_pipeline.predict(
|
||||
return self.paddlex_pipeline.predict(
|
||||
input,
|
||||
use_doc_orientation_classify=use_doc_orientation_classify,
|
||||
use_doc_unwarping=use_doc_unwarping,
|
||||
use_textline_orientation=use_textline_orientation,
|
||||
use_general_ocr=use_general_ocr,
|
||||
use_seal_recognition=use_seal_recognition,
|
||||
use_table_recognition=use_table_recognition,
|
||||
use_formula_recognition=use_formula_recognition,
|
||||
use_chart_recognition=use_chart_recognition,
|
||||
use_region_detection=use_region_detection,
|
||||
layout_threshold=layout_threshold,
|
||||
layout_nms=layout_nms,
|
||||
layout_unclip_ratio=layout_unclip_ratio,
|
||||
@ -149,13 +159,86 @@ class PPStructureV3(PaddleXPipelineWrapper):
|
||||
seal_det_box_thresh=seal_det_box_thresh,
|
||||
seal_det_unclip_ratio=seal_det_unclip_ratio,
|
||||
seal_rec_score_thresh=seal_rec_score_thresh,
|
||||
use_table_cells_ocr_results=use_table_cells_ocr_results,
|
||||
use_wired_table_cells_trans_to_html=use_wired_table_cells_trans_to_html,
|
||||
use_wireless_table_cells_trans_to_html=use_wireless_table_cells_trans_to_html,
|
||||
use_table_orientation_classify=use_table_orientation_classify,
|
||||
use_ocr_results_with_table_cells=use_ocr_results_with_table_cells,
|
||||
use_e2e_wired_table_rec_model=use_e2e_wired_table_rec_model,
|
||||
use_e2e_wireless_table_rec_model=use_e2e_wireless_table_rec_model,
|
||||
**kwargs,
|
||||
):
|
||||
result.append(res)
|
||||
return result
|
||||
)
|
||||
|
||||
def predict(
|
||||
self,
|
||||
input,
|
||||
use_doc_orientation_classify=False,
|
||||
use_doc_unwarping=False,
|
||||
use_textline_orientation=None,
|
||||
use_seal_recognition=None,
|
||||
use_table_recognition=None,
|
||||
use_formula_recognition=None,
|
||||
use_chart_recognition=False,
|
||||
use_region_detection=None,
|
||||
layout_threshold=None,
|
||||
layout_nms=None,
|
||||
layout_unclip_ratio=None,
|
||||
layout_merge_bboxes_mode=None,
|
||||
text_det_limit_side_len=None,
|
||||
text_det_limit_type=None,
|
||||
text_det_thresh=None,
|
||||
text_det_box_thresh=None,
|
||||
text_det_unclip_ratio=None,
|
||||
text_rec_score_thresh=None,
|
||||
seal_det_limit_side_len=None,
|
||||
seal_det_limit_type=None,
|
||||
seal_det_thresh=None,
|
||||
seal_det_box_thresh=None,
|
||||
seal_det_unclip_ratio=None,
|
||||
seal_rec_score_thresh=None,
|
||||
use_wired_table_cells_trans_to_html=False,
|
||||
use_wireless_table_cells_trans_to_html=False,
|
||||
use_table_orientation_classify=True,
|
||||
use_ocr_results_with_table_cells=True,
|
||||
use_e2e_wired_table_rec_model=False,
|
||||
use_e2e_wireless_table_rec_model=True,
|
||||
**kwargs,
|
||||
):
|
||||
return list(
|
||||
self.predict_iter(
|
||||
input,
|
||||
use_doc_orientation_classify=use_doc_orientation_classify,
|
||||
use_doc_unwarping=use_doc_unwarping,
|
||||
use_textline_orientation=use_textline_orientation,
|
||||
use_seal_recognition=use_seal_recognition,
|
||||
use_table_recognition=use_table_recognition,
|
||||
use_formula_recognition=use_formula_recognition,
|
||||
use_chart_recognition=use_chart_recognition,
|
||||
use_region_detection=use_region_detection,
|
||||
layout_threshold=layout_threshold,
|
||||
layout_nms=layout_nms,
|
||||
layout_unclip_ratio=layout_unclip_ratio,
|
||||
layout_merge_bboxes_mode=layout_merge_bboxes_mode,
|
||||
text_det_limit_side_len=text_det_limit_side_len,
|
||||
text_det_limit_type=text_det_limit_type,
|
||||
text_det_thresh=text_det_thresh,
|
||||
text_det_box_thresh=text_det_box_thresh,
|
||||
text_det_unclip_ratio=text_det_unclip_ratio,
|
||||
text_rec_score_thresh=text_rec_score_thresh,
|
||||
seal_det_limit_side_len=seal_det_limit_side_len,
|
||||
seal_det_limit_type=seal_det_limit_type,
|
||||
seal_det_thresh=seal_det_thresh,
|
||||
seal_det_box_thresh=seal_det_box_thresh,
|
||||
seal_det_unclip_ratio=seal_det_unclip_ratio,
|
||||
seal_rec_score_thresh=seal_rec_score_thresh,
|
||||
use_wired_table_cells_trans_to_html=use_wired_table_cells_trans_to_html,
|
||||
use_wireless_table_cells_trans_to_html=use_wireless_table_cells_trans_to_html,
|
||||
use_table_orientation_classify=use_table_orientation_classify,
|
||||
use_ocr_results_with_table_cells=use_ocr_results_with_table_cells,
|
||||
use_e2e_wired_table_rec_model=use_e2e_wired_table_rec_model,
|
||||
use_e2e_wireless_table_rec_model=use_e2e_wireless_table_rec_model,
|
||||
**kwargs,
|
||||
)
|
||||
)
|
||||
|
||||
@classmethod
|
||||
def get_cli_subcommand_executor(cls):
|
||||
@ -169,7 +252,6 @@ class PPStructureV3(PaddleXPipelineWrapper):
|
||||
"SubPipelines.DocPreprocessor.use_doc_unwarping": self._params[
|
||||
"use_doc_unwarping"
|
||||
],
|
||||
"use_general_ocr": self._params["use_general_ocr"],
|
||||
"use_seal_recognition": self._params["use_seal_recognition"],
|
||||
"use_table_recognition": self._params["use_table_recognition"],
|
||||
"use_formula_recognition": self._params["use_formula_recognition"],
|
||||
@ -187,6 +269,21 @@ class PPStructureV3(PaddleXPipelineWrapper):
|
||||
"SubModules.LayoutDetection.layout_merge_bboxes_mode": self._params[
|
||||
"layout_merge_bboxes_mode"
|
||||
],
|
||||
"SubModules.ChartRecognition.model_name": self._params[
|
||||
"chart_recognition_model_name"
|
||||
],
|
||||
"SubModules.ChartRecognition.model_dir": self._params[
|
||||
"chart_recognition_model_dir"
|
||||
],
|
||||
"SubModules.ChartRecognition.batch_size": self._params[
|
||||
"chart_recognition_batch_size"
|
||||
],
|
||||
"SubModules.RegionDetection.model_name": self._params[
|
||||
"region_detection_model_name"
|
||||
],
|
||||
"SubModules.RegionDetection.model_dir": self._params[
|
||||
"region_detection_model_dir"
|
||||
],
|
||||
"SubPipelines.DocPreprocessor.SubModules.DocOrientationClassify.model_name": self._params[
|
||||
"doc_orientation_classify_model_name"
|
||||
],
|
||||
@ -353,6 +450,33 @@ class PPStructureV3CLISubcommandExecutor(PipelineCLISubcommandExecutor):
|
||||
help="Overlapping box filtering method.",
|
||||
)
|
||||
|
||||
subparser.add_argument(
|
||||
"--chart_recognition_model_name",
|
||||
type=str,
|
||||
help="Name of the chart recognition model.",
|
||||
)
|
||||
subparser.add_argument(
|
||||
"--chart_recognition_model_dir",
|
||||
type=str,
|
||||
help="Path to the chart recognition model directory.",
|
||||
)
|
||||
subparser.add_argument(
|
||||
"--chart_recognition_batch_size",
|
||||
type=int,
|
||||
help="Batch size for the chart recognition model.",
|
||||
)
|
||||
|
||||
subparser.add_argument(
|
||||
"--region_detection_model_name",
|
||||
type=str,
|
||||
help="Name of the region detection model.",
|
||||
)
|
||||
subparser.add_argument(
|
||||
"--region_detection_model_dir",
|
||||
type=str,
|
||||
help="Path to the region detection model directory.",
|
||||
)
|
||||
|
||||
subparser.add_argument(
|
||||
"--doc_orientation_classify_model_name",
|
||||
type=str,
|
||||
@ -571,18 +695,15 @@ class PPStructureV3CLISubcommandExecutor(PipelineCLISubcommandExecutor):
|
||||
subparser.add_argument(
|
||||
"--use_doc_orientation_classify",
|
||||
type=str2bool,
|
||||
default=False,
|
||||
help="Whether to use the document image orientation classification model.",
|
||||
)
|
||||
subparser.add_argument(
|
||||
"--use_doc_unwarping",
|
||||
type=str2bool,
|
||||
default=False,
|
||||
help="Whether to use the text image unwarping model.",
|
||||
)
|
||||
subparser.add_argument(
|
||||
"--use_general_ocr",
|
||||
type=str2bool,
|
||||
help="Whether to use general OCR.",
|
||||
)
|
||||
subparser.add_argument(
|
||||
"--use_seal_recognition",
|
||||
type=str2bool,
|
||||
@ -598,6 +719,17 @@ class PPStructureV3CLISubcommandExecutor(PipelineCLISubcommandExecutor):
|
||||
type=str2bool,
|
||||
help="Whether to use formula recognition.",
|
||||
)
|
||||
subparser.add_argument(
|
||||
"--use_chart_recognition",
|
||||
type=str2bool,
|
||||
default=False,
|
||||
help="Whether to use chart recognition.",
|
||||
)
|
||||
subparser.add_argument(
|
||||
"--use_region_detection",
|
||||
type=str2bool,
|
||||
help="Whether to use region detection.",
|
||||
)
|
||||
|
||||
def execute_with_args(self, args):
|
||||
params = get_subcommand_args(args)
|
||||
|
||||
@ -12,7 +12,7 @@
|
||||
# See the License for the specific language governing permissions and
|
||||
# limitations under the License.
|
||||
|
||||
from ..utils.cli import (
|
||||
from .._utils.cli import (
|
||||
add_simple_inference_args,
|
||||
get_subcommand_args,
|
||||
perform_simple_inference,
|
||||
@ -84,6 +84,45 @@ class SealRecognition(PaddleXPipelineWrapper):
|
||||
def _paddlex_pipeline_name(self):
|
||||
return "seal_recognition"
|
||||
|
||||
def predict_iter(
|
||||
self,
|
||||
input,
|
||||
*,
|
||||
use_doc_orientation_classify=None,
|
||||
use_doc_unwarping=None,
|
||||
use_layout_detection=None,
|
||||
layout_det_res=None,
|
||||
layout_threshold=None,
|
||||
layout_nms=None,
|
||||
layout_unclip_ratio=None,
|
||||
layout_merge_bboxes_mode=None,
|
||||
seal_det_limit_side_len=None,
|
||||
seal_det_limit_type=None,
|
||||
seal_det_thresh=None,
|
||||
seal_det_box_thresh=None,
|
||||
seal_det_unclip_ratio=None,
|
||||
seal_rec_score_thresh=None,
|
||||
**kwargs,
|
||||
):
|
||||
return self.paddlex_pipeline.predict(
|
||||
input,
|
||||
use_doc_orientation_classify=use_doc_orientation_classify,
|
||||
use_doc_unwarping=use_doc_unwarping,
|
||||
use_layout_detection=use_layout_detection,
|
||||
layout_det_res=layout_det_res,
|
||||
layout_threshold=layout_threshold,
|
||||
layout_nms=layout_nms,
|
||||
layout_unclip_ratio=layout_unclip_ratio,
|
||||
layout_merge_bboxes_mode=layout_merge_bboxes_mode,
|
||||
seal_det_limit_side_len=seal_det_limit_side_len,
|
||||
seal_det_limit_type=seal_det_limit_type,
|
||||
seal_det_thresh=seal_det_thresh,
|
||||
seal_det_box_thresh=seal_det_box_thresh,
|
||||
seal_det_unclip_ratio=seal_det_unclip_ratio,
|
||||
seal_rec_score_thresh=seal_rec_score_thresh,
|
||||
**kwargs,
|
||||
)
|
||||
|
||||
def predict(
|
||||
self,
|
||||
input,
|
||||
@ -104,27 +143,26 @@ class SealRecognition(PaddleXPipelineWrapper):
|
||||
seal_rec_score_thresh=None,
|
||||
**kwargs,
|
||||
):
|
||||
result = []
|
||||
for res in self.paddlex_pipeline.predict(
|
||||
input,
|
||||
use_doc_orientation_classify=use_doc_orientation_classify,
|
||||
use_doc_unwarping=use_doc_unwarping,
|
||||
use_layout_detection=use_layout_detection,
|
||||
layout_det_res=layout_det_res,
|
||||
layout_threshold=layout_threshold,
|
||||
layout_nms=layout_nms,
|
||||
layout_unclip_ratio=layout_unclip_ratio,
|
||||
layout_merge_bboxes_mode=layout_merge_bboxes_mode,
|
||||
seal_det_limit_side_len=seal_det_limit_side_len,
|
||||
seal_det_limit_type=seal_det_limit_type,
|
||||
seal_det_thresh=seal_det_thresh,
|
||||
seal_det_box_thresh=seal_det_box_thresh,
|
||||
seal_det_unclip_ratio=seal_det_unclip_ratio,
|
||||
seal_rec_score_thresh=seal_rec_score_thresh,
|
||||
**kwargs,
|
||||
):
|
||||
result.append(res)
|
||||
return result
|
||||
return list(
|
||||
self.predict_iter(
|
||||
input,
|
||||
use_doc_orientation_classify=use_doc_orientation_classify,
|
||||
use_doc_unwarping=use_doc_unwarping,
|
||||
use_layout_detection=use_layout_detection,
|
||||
layout_det_res=layout_det_res,
|
||||
layout_threshold=layout_threshold,
|
||||
layout_nms=layout_nms,
|
||||
layout_unclip_ratio=layout_unclip_ratio,
|
||||
layout_merge_bboxes_mode=layout_merge_bboxes_mode,
|
||||
seal_det_limit_side_len=seal_det_limit_side_len,
|
||||
seal_det_limit_type=seal_det_limit_type,
|
||||
seal_det_thresh=seal_det_thresh,
|
||||
seal_det_box_thresh=seal_det_box_thresh,
|
||||
seal_det_unclip_ratio=seal_det_unclip_ratio,
|
||||
seal_rec_score_thresh=seal_rec_score_thresh,
|
||||
**kwargs,
|
||||
)
|
||||
)
|
||||
|
||||
@classmethod
|
||||
def get_cli_subcommand_executor(cls):
|
||||
|
||||
@ -12,7 +12,7 @@
|
||||
# See the License for the specific language governing permissions and
|
||||
# limitations under the License.
|
||||
|
||||
from ..utils.cli import (
|
||||
from .._utils.cli import (
|
||||
add_simple_inference_args,
|
||||
get_subcommand_args,
|
||||
perform_simple_inference,
|
||||
@ -69,6 +69,52 @@ class TableRecognitionPipelineV2(PaddleXPipelineWrapper):
|
||||
def _paddlex_pipeline_name(self):
|
||||
return "table_recognition_v2"
|
||||
|
||||
def predict_iter(
|
||||
self,
|
||||
input,
|
||||
use_doc_orientation_classify=None,
|
||||
use_doc_unwarping=None,
|
||||
use_layout_detection=None,
|
||||
use_ocr_model=None,
|
||||
overall_ocr_res=None,
|
||||
layout_det_res=None,
|
||||
text_det_limit_side_len=None,
|
||||
text_det_limit_type=None,
|
||||
text_det_thresh=None,
|
||||
text_det_box_thresh=None,
|
||||
text_det_unclip_ratio=None,
|
||||
text_rec_score_thresh=None,
|
||||
use_e2e_wired_table_rec_model=False,
|
||||
use_e2e_wireless_table_rec_model=False,
|
||||
use_wired_table_cells_trans_to_html=False,
|
||||
use_wireless_table_cells_trans_to_html=False,
|
||||
use_table_orientation_classify=True,
|
||||
use_ocr_results_with_table_cells=True,
|
||||
**kwargs,
|
||||
):
|
||||
return self.paddlex_pipeline.predict(
|
||||
input,
|
||||
use_doc_orientation_classify=use_doc_orientation_classify,
|
||||
use_doc_unwarping=use_doc_unwarping,
|
||||
use_layout_detection=use_layout_detection,
|
||||
use_ocr_model=use_ocr_model,
|
||||
overall_ocr_res=overall_ocr_res,
|
||||
layout_det_res=layout_det_res,
|
||||
text_det_limit_side_len=text_det_limit_side_len,
|
||||
text_det_limit_type=text_det_limit_type,
|
||||
text_det_thresh=text_det_thresh,
|
||||
text_det_box_thresh=text_det_box_thresh,
|
||||
text_det_unclip_ratio=text_det_unclip_ratio,
|
||||
text_rec_score_thresh=text_rec_score_thresh,
|
||||
use_e2e_wired_table_rec_model=use_e2e_wired_table_rec_model,
|
||||
use_e2e_wireless_table_rec_model=use_e2e_wireless_table_rec_model,
|
||||
use_wired_table_cells_trans_to_html=use_wired_table_cells_trans_to_html,
|
||||
use_wireless_table_cells_trans_to_html=use_wireless_table_cells_trans_to_html,
|
||||
use_table_orientation_classify=use_table_orientation_classify,
|
||||
use_ocr_results_with_table_cells=use_ocr_results_with_table_cells,
|
||||
**kwargs,
|
||||
)
|
||||
|
||||
def predict(
|
||||
self,
|
||||
input,
|
||||
@ -84,33 +130,38 @@ class TableRecognitionPipelineV2(PaddleXPipelineWrapper):
|
||||
text_det_box_thresh=None,
|
||||
text_det_unclip_ratio=None,
|
||||
text_rec_score_thresh=None,
|
||||
use_table_cells_ocr_results=None,
|
||||
use_e2e_wired_table_rec_model=None,
|
||||
use_e2e_wireless_table_rec_model=None,
|
||||
use_e2e_wired_table_rec_model=False,
|
||||
use_e2e_wireless_table_rec_model=False,
|
||||
use_wired_table_cells_trans_to_html=False,
|
||||
use_wireless_table_cells_trans_to_html=False,
|
||||
use_table_orientation_classify=True,
|
||||
use_ocr_results_with_table_cells=True,
|
||||
**kwargs,
|
||||
):
|
||||
result = []
|
||||
for res in self.paddlex_pipeline.predict(
|
||||
input,
|
||||
use_doc_orientation_classify=use_doc_orientation_classify,
|
||||
use_doc_unwarping=use_doc_unwarping,
|
||||
use_layout_detection=use_layout_detection,
|
||||
use_ocr_model=use_ocr_model,
|
||||
overall_ocr_res=overall_ocr_res,
|
||||
layout_det_res=layout_det_res,
|
||||
text_det_limit_side_len=text_det_limit_side_len,
|
||||
text_det_limit_type=text_det_limit_type,
|
||||
text_det_thresh=text_det_thresh,
|
||||
text_det_box_thresh=text_det_box_thresh,
|
||||
text_det_unclip_ratio=text_det_unclip_ratio,
|
||||
text_rec_score_thresh=text_rec_score_thresh,
|
||||
use_table_cells_ocr_results=use_table_cells_ocr_results,
|
||||
use_e2e_wired_table_rec_model=use_e2e_wired_table_rec_model,
|
||||
use_e2e_wireless_table_rec_model=use_e2e_wireless_table_rec_model,
|
||||
**kwargs,
|
||||
):
|
||||
result.append(res)
|
||||
return result
|
||||
return list(
|
||||
self.predict_iter(
|
||||
input,
|
||||
use_doc_orientation_classify=use_doc_orientation_classify,
|
||||
use_doc_unwarping=use_doc_unwarping,
|
||||
use_layout_detection=use_layout_detection,
|
||||
use_ocr_model=use_ocr_model,
|
||||
overall_ocr_res=overall_ocr_res,
|
||||
layout_det_res=layout_det_res,
|
||||
text_det_limit_side_len=text_det_limit_side_len,
|
||||
text_det_limit_type=text_det_limit_type,
|
||||
text_det_thresh=text_det_thresh,
|
||||
text_det_box_thresh=text_det_box_thresh,
|
||||
text_det_unclip_ratio=text_det_unclip_ratio,
|
||||
text_rec_score_thresh=text_rec_score_thresh,
|
||||
use_e2e_wired_table_rec_model=use_e2e_wired_table_rec_model,
|
||||
use_e2e_wireless_table_rec_model=use_e2e_wireless_table_rec_model,
|
||||
use_wired_table_cells_trans_to_html=use_wired_table_cells_trans_to_html,
|
||||
use_wireless_table_cells_trans_to_html=use_wireless_table_cells_trans_to_html,
|
||||
use_table_orientation_classify=use_table_orientation_classify,
|
||||
use_ocr_results_with_table_cells=use_ocr_results_with_table_cells,
|
||||
**kwargs,
|
||||
)
|
||||
)
|
||||
|
||||
@classmethod
|
||||
def get_cli_subcommand_executor(cls):
|
||||
|
||||
@ -12,6 +12,10 @@
|
||||
# See the License for the specific language governing permissions and
|
||||
# limitations under the License.
|
||||
|
||||
import time
|
||||
|
||||
from .logging import logger
|
||||
|
||||
|
||||
def str2bool(v, /):
|
||||
return v.lower() in ("true", "yes", "t", "y", "1")
|
||||
@ -24,9 +28,15 @@ def get_subcommand_args(args):
|
||||
return args
|
||||
|
||||
|
||||
def add_simple_inference_args(subparser):
|
||||
def add_simple_inference_args(subparser, *, input_help=None):
|
||||
if input_help is None:
|
||||
input_help = "Input path or URL."
|
||||
subparser.add_argument(
|
||||
"-i", "--input", type=str, required=True, help="Input path or URL."
|
||||
"-i",
|
||||
"--input",
|
||||
type=str,
|
||||
required=True,
|
||||
help=input_help,
|
||||
)
|
||||
subparser.add_argument(
|
||||
"--save_path",
|
||||
@ -42,9 +52,12 @@ def perform_simple_inference(wrapper_cls, params):
|
||||
|
||||
wrapper = wrapper_cls(**params)
|
||||
|
||||
result = wrapper.predict(input_)
|
||||
result = wrapper.predict_iter(input_)
|
||||
|
||||
for res in result:
|
||||
t1 = time.time()
|
||||
for i, res in enumerate(result):
|
||||
logger.info(f"Processed item {i} in {(time.time()-t1) * 1000} ms")
|
||||
t1 = time.time()
|
||||
res.print()
|
||||
if save_path:
|
||||
res.save_all(save_path)
|
||||
@ -33,7 +33,6 @@ def test_visual_predict(pp_chatocrv4_doc_pipeline, image_path):
|
||||
[
|
||||
{"use_doc_orientation_classify": False},
|
||||
{"use_doc_unwarping": False},
|
||||
{"use_general_ocr": False},
|
||||
{"use_table_recognition": False},
|
||||
{"layout_threshold": 0.88},
|
||||
{"layout_threshold": [0.45, 0.4]},
|
||||
|
||||
@ -36,7 +36,6 @@ def test_visual_predict(pp_structurev3_pipeline, image_path):
|
||||
[
|
||||
{"use_doc_orientation_classify": False},
|
||||
{"use_doc_unwarping": False},
|
||||
{"use_general_ocr": False},
|
||||
{"use_table_recognition": False},
|
||||
{"use_formula_recognition": False},
|
||||
{"layout_threshold": 0.88},
|
||||
|
||||
Loading…
x
Reference in New Issue
Block a user