Skip to content

📄 Awesome OCR multiple programing languages toolkits based on ONNX Runtime, OpenVINO, MNN, PaddlePaddle, TensorRT and PyTorch.

Update README

SWHLabbcea65b8b5
diff --git a/README_en.md b/README_en.md
index d8bb223..2ad0be0 100644
--- a/README_en.md
+++ b/README_en.md
@@ -22,7 +22,8 @@
- [RapidOCR](#rapidocr)
- [Introduction](#introduction)
- [Recently updates](#recently-updates)
- - [🔉2022-02-24 update](#2022-02-24-update)
+ - [🎧2022-04-04 update](#2022-04-04-update)
+ - [2022-02-24 update](#2022-02-24-update)
- [2021-12-18 update](#2021-12-18-update)
- [2021-11-28 update](#2021-11-28-update)
- [2021-11-13 update](#2021-11-13-update)
@@ -65,7 +66,12 @@
- Based on Baidu's open source PaddleOCR model and training, anyone can use this inference library, or use Baidu's paddlepaddle framework for model optimization according to their own needs.
## Recently updates
-#### 🔉2022-02-24 update
+#### 🎧2022-04-04 update
+- Add suport for OpenVINO under python
+- Give the performance comparison table of OpenVINO and ONNXRuntime
+- For details:[python/README](./python/README.md)
+
+#### 2022-02-24 update
- Optimize the inference code of the python part.
- Add inference code examples that use the different language models.
- For details, see: [python/README](./python/README.md)
diff --git a/commonlib/README.md b/commonlib/README.md
deleted file mode 100755
index b3a4252..0000000
--- a/commonlib/README.md
+++ /dev/null
@@ -1 +0,0 @@
-placeholder
\ No newline at end of file
diff --git a/datasets/README.md b/datasets/README.md
deleted file mode 100755
index 93e1fd9..0000000
--- a/datasets/README.md
+++ /dev/null
@@ -1,9 +0,0 @@
-## 简单汇总OCR相关数据集
-#### [From PaddleOCR: 通用中英文OCR数据集](https://github.com/PaddlePaddle/PaddleOCR/blob/release/2.1/doc/doc_ch/datasets.md)
-
-#### [From PaddleOCR: 手写OCR数据集](https://github.com/PaddlePaddle/PaddleOCR/blob/release/2.1/doc/doc_ch/handwritten_datasets.md)
-
-#### [From PaddleOCR: 垂类多语言OCR数据集](https://github.com/PaddlePaddle/PaddleOCR/blob/release/2.1/doc/doc_ch/vertical_and_multilingual_datasets.md)
-
-#### [WenmuZhou/OCR_DataSet](https://github.com/WenmuZhou/OCR_DataSet)
-
diff --git a/python/onnxruntime_infer/README.md b/python/onnxruntime_infer/README.md
index 22b78cf..2257b4d 100644
--- a/python/onnxruntime_infer/README.md
+++ b/python/onnxruntime_infer/README.md
@@ -1,5 +1,6 @@
### 使用步骤
-1. 下载相应模型和用于显示的字体文件
+1. 下载`python/onnxruntime_infer`目录到本地
+2. 下载相应模型和用于显示的字体文件
- [百度网盘 提取码:30jv](https://pan.baidu.com/s/1qkqWK4wRdMjqGGbzR-FyWg) | [Google Drive](https://drive.google.com/drive/folders/1x_a9KpCo_1blxH1xFOfgKVkw1HYRVywY?usp=sharing)
- 下载之后模型和相应字体文件放在`fonts`和`models`下,最终目录结构如下:
@@ -19,7 +20,7 @@
`-- korean.ttf
```
-2. 运行
+3. 运行
- 接口调用方式运行
- 如果想要使用其他语言的识别模型,可以只对`rec_model_path`和`keys_path`作对应更改即可
- **!!!各个不同语言的推理代码区别只在于模型和字典文件!!!**
@@ -105,7 +106,7 @@
### onnxruntime-gpu版推理配置
-1. ONNXRuntime-gpu需要严格按照与cuda、cudnn版本对应来安装,具体参考[文档](https://onnxruntime.ai/docs/execution-providers/CUDA-ExecutionProvider.html#requirements), **这一步关乎后面是否可以成功调用GPU**
+1. onnxruntime-gpu需要严格按照与cuda、cudnn版本对应来安装,具体参考[文档](https://onnxruntime.ai/docs/execution-providers/CUDA-ExecutionProvider.html#requirements),**这一步关乎后面是否可以成功调用GPU**
2. 推理代码中,加载onnx模型部分,用以下对应语言代码替换即可,详细参见:[官方教程](https://onnxruntime.ai/docs/execution-providers/CUDA-ExecutionProvider.html)
- python版本
```python
@@ -139,7 +140,7 @@
SessionOptionsAppendExecutionProvider_CUDA(session_options, &options);
```
-3. 推理时间粗略对比(跑完RapidOCR整个项目)
+3. 推理时间粗略对比(完整跑完一张图像)
|推理方式|推理图像数目|耗费时间|
|:---:|:---:|:---:|