mirror of
https://github.com/PaddlePaddle/FastDeploy.git
synced 2025-10-05 16:48:03 +08:00
Add PaddleOCRv3 & PaddleOCRv2 Support (#139)
* Add PaddleOCR Support * Add PaddleOCR Support * Add PaddleOCRv3 Support * Add PaddleOCRv3 Support * Update README.md * Update README.md * Update README.md * Update README.md * Add PaddleOCRv3 Support * Add PaddleOCRv3 Supports * Add PaddleOCRv3 Suport * Fix Rec diff * Remove useless functions * Remove useless comments * Add PaddleOCRv2 Support
This commit is contained in:
yunyaoXYY
14
examples/vision/ocr/PPOCRSystemv2/cpp/CMakeLists.txt
Normal file
14
examples/vision/ocr/PPOCRSystemv2/cpp/CMakeLists.txt
Normal file
@@ -0,0 +1,14 @@
|
||||
PROJECT(infer_demo C CXX)
|
||||
CMAKE_MINIMUM_REQUIRED (VERSION 3.12)
|
||||
|
||||
# 指定下载解压后的fastdeploy库路径
|
||||
option(FASTDEPLOY_INSTALL_DIR "Path of downloaded fastdeploy sdk.")
|
||||
|
||||
include(${FASTDEPLOY_INSTALL_DIR}/FastDeploy.cmake)
|
||||
|
||||
# 添加FastDeploy依赖头文件
|
||||
include_directories(${FASTDEPLOY_INCS})
|
||||
|
||||
add_executable(infer_demo ${PROJECT_SOURCE_DIR}/infer.cc)
|
||||
# 添加FastDeploy库依赖
|
||||
target_link_libraries(infer_demo ${FASTDEPLOY_LIBS})
|
||||
139
examples/vision/ocr/PPOCRSystemv2/cpp/README.md
Normal file
139
examples/vision/ocr/PPOCRSystemv2/cpp/README.md
Normal file
@@ -0,0 +1,139 @@
|
||||
# PPOCRSystemv2 C++部署示例
|
||||
|
||||
本目录下提供`infer.cc`快速完成PPOCRSystemv2在CPU/GPU，以及GPU上通过TensorRT加速部署的示例。
|
||||
|
||||
在部署前，需确认以下两个步骤
|
||||
|
||||
- 1. 软硬件环境满足要求，参考[FastDeploy环境要求](../../../../../docs/the%20software%20and%20hardware%20requirements.md)
|
||||
- 2. 根据开发环境，下载预编译部署库和samples代码，参考[FastDeploy预编译库](../../../../../docs/quick_start)
|
||||
|
||||
以Linux上CPU推理为例，在本目录执行如下命令即可完成编译测试
|
||||
|
||||
```
|
||||
mkdir build
|
||||
cd build
|
||||
wget https://https://bj.bcebos.com/paddlehub/fastdeploy/cpp/fastdeploy-linux-x64-gpu-0.2.0.tgz
|
||||
tar xvf fastdeploy-linux-x64-0.2.0.tgz
|
||||
cmake .. -DFASTDEPLOY_INSTALL_DIR=${PWD}/fastdeploy-linux-x64-0.2.0
|
||||
make -j
|
||||
|
||||
|
||||
# 下载模型,图片和label文件
|
||||
wget https://paddleocr.bj.bcebos.com/PP-OCRv2/chinese/ch_PP-OCRv2_det_infer.tar
|
||||
tar xvf ch_PP-OCRv2_det_infer.tar
|
||||
|
||||
wget https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_cls_infer.tar
|
||||
tar xvf ch_ppocr_mobile_v2.0_cls_infer.tar
|
||||
|
||||
wget https://paddleocr.bj.bcebos.com/PP-OCRv2/chinese/ch_PP-OCRv2_rec_infer.tar
|
||||
tar xvf ch_PP-OCRv2_rec_infer.tar
|
||||
|
||||
wget https://raw.githubusercontent.com/PaddlePaddle/PaddleOCR/release/2.5/doc/imgs/12.jpg
|
||||
|
||||
wget https://raw.githubusercontent.com/PaddlePaddle/PaddleOCR/release/2.5/ppocr/utils/ppocr_keys_v1.txt
|
||||
|
||||
|
||||
# CPU推理
|
||||
./infer_demo ./ch_PP-OCRv2_det_infer ./ch_ppocr_mobile_v2.0_cls_infer ./ch_PP-OCRv2_rec_infer ./ppocr_keys_v1.txt ./12.jpg 0
|
||||
# GPU推理
|
||||
./infer_demo ./ch_PP-OCRv2_det_infer ./ch_ppocr_mobile_v2.0_cls_infer ./ch_PP-OCRv2_rec_infer ./ppocr_keys_v1.txt ./12.jpg 1
|
||||
# GPU上TensorRT推理
|
||||
./infer_demo ./ch_PP-OCRv2_det_infer ./ch_ppocr_mobile_v2.0_cls_infer ./ch_PP-OCRv2_rec_infer ./ppocr_keys_v1.txt ./12.jpg 2
|
||||
# OCR还支持det/cls/rec三个模型的组合使用，例如当我们不想使用cls模型的时候，只需要给cls模型路径的位置，传入一个空的字符串, 例子如下
|
||||
./infer_demo ./ch_PP-OCRv2_det_infer "" ./ch_PP-OCRv2_rec_infer ./ppocr_keys_v1.txt ./12.jpg 0
|
||||
```
|
||||
|
||||
运行完成可视化结果如下图所示
|
||||
|
||||
<img width="640" src="https://user-images.githubusercontent.com/109218879/185826024-f7593a0c-1bd2-4a60-b76c-15588484fa08.jpg">
|
||||
|
||||
|
||||
## PPOCRSystemv2 C++接口
|
||||
|
||||
### PPOCRSystemv2类
|
||||
|
||||
```
|
||||
fastdeploy::application::ocrsystem::PPOCRSystemv2(fastdeploy::vision::ocr::DBDetector* ocr_det = nullptr,
|
||||
fastdeploy::vision::ocr::Classifier* ocr_cls = nullptr,
|
||||
fastdeploy::vision::ocr::Recognizer* ocr_rec = nullptr);
|
||||
```
|
||||
|
||||
PPOCRSystemv2 的初始化，由检测，分类和识别模型串联构成
|
||||
|
||||
**参数**
|
||||
|
||||
> * **DBDetector**(model): OCR中的检测模型
|
||||
> * **Classifier**(model): OCR中的分类模型
|
||||
> * **Recognizer**(model): OCR中的识别模型
|
||||
|
||||
#### Predict函数
|
||||
|
||||
> ```
|
||||
> std::vector<std::vector<fastdeploy::vision::OCRResult>> ocr_results =
|
||||
> PPOCRSystemv2.Predict(std::vector<cv::Mat> cv_all_imgs);
|
||||
>
|
||||
> ```
|
||||
>
|
||||
> 模型预测接口，输入一个可装入多张图片的图片列表，后可输出检测结果。
|
||||
>
|
||||
> **参数**
|
||||
>
|
||||
> > * **cv_all_imgs**: 输入图像，注意需为HWC，BGR格式
|
||||
> > * **ocr_results**: OCR结果,包括由检测模型输出的检测框位置,分类模型输出的方向分类,以及识别模型输出的识别结果, OCRResult说明参考[视觉模型预测结果](../../../../../docs/api/vision_results/)
|
||||
|
||||
|
||||
## DBDetector C++接口
|
||||
|
||||
### DBDetector类
|
||||
|
||||
```
|
||||
fastdeploy::vision::ocr::DBDetector(const std::string& model_file, const std::string& params_file = "",
|
||||
const RuntimeOption& custom_option = RuntimeOption(),
|
||||
const Frontend& model_format = Frontend::PADDLE);
|
||||
```
|
||||
|
||||
DBDetector模型加载和初始化，其中模型为paddle模型格式。
|
||||
|
||||
**参数**
|
||||
|
||||
> * **model_file**(str): 模型文件路径
|
||||
> * **params_file**(str): 参数文件路径，当模型格式为ONNX时，此参数传入空字符串即可
|
||||
> * **runtime_option**(RuntimeOption): 后端推理配置，默认为None，即采用默认配置
|
||||
> * **model_format**(Frontend): 模型格式，默认为Paddle格式
|
||||
|
||||
### Classifier类与DBDetector类相同
|
||||
|
||||
### Recognizer类
|
||||
```
|
||||
Recognizer(const std::string& model_file,
|
||||
const std::string& params_file = "",
|
||||
const std::string& label_path = "",
|
||||
const RuntimeOption& custom_option = RuntimeOption(),
|
||||
const Frontend& model_format = Frontend::PADDLE);
|
||||
```
|
||||
Recognizer类初始化时,需要在label_path参数中,输入识别模型所需的label文件，其他参数均与DBDetector类相同
|
||||
|
||||
**参数**
|
||||
> * **label_path**(str): 识别模型的label文件路径
|
||||
|
||||
|
||||
### 类成员变量
|
||||
#### DBDetector预处理参数
|
||||
用户可按照自己的实际需求，修改下列预处理参数，从而影响最终的推理和部署效果
|
||||
|
||||
> > * **max_side_len**(int): 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,默认为960
|
||||
> > * **det_db_thresh**(double): DB模型输出预测图的二值化阈值，默认为0.3
|
||||
> > * **det_db_box_thresh**(double): DB模型输出框的阈值，低于此值的预测框会被丢弃，默认为0.6
|
||||
> > * **det_db_unclip_ratio**(double): DB模型输出框扩大的比例，默认为1.5
|
||||
> > * **det_db_score_mode**(string):DB后处理中计算文本框平均得分的方式,默认为slow，即求polygon区域的平均分数的方式
|
||||
> > * **use_dilation**(bool):是否对检测输出的feature map做膨胀处理,默认为Fasle
|
||||
|
||||
#### Classifier预处理参数
|
||||
用户可按照自己的实际需求，修改下列预处理参数，从而影响最终的推理和部署效果
|
||||
|
||||
> > * **cls_thresh**(double): 当分类模型输出的得分超过此阈值，输入的图片将被翻转，默认为0.9
|
||||
|
||||
|
||||
- [模型介绍](../../)
|
||||
- [Python部署](../python)
|
||||
- [视觉模型预测结果](../../../../../docs/api/vision_results/)
|
||||
295
examples/vision/ocr/PPOCRSystemv2/cpp/infer.cc
Normal file
295
examples/vision/ocr/PPOCRSystemv2/cpp/infer.cc
Normal file
@@ -0,0 +1,295 @@
|
||||
// Copyright (c) 2022 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.
|
||||
|
||||
#include "fastdeploy/vision.h"
|
||||
#ifdef WIN32
|
||||
const char sep = '\\';
|
||||
#else
|
||||
const char sep = '/';
|
||||
#endif
|
||||
|
||||
void CpuInfer(const std::string& det_model_dir,
|
||||
const std::string& cls_model_dir,
|
||||
const std::string& rec_model_dir,
|
||||
const std::string& rec_label_file,
|
||||
const std::string& image_file) {
|
||||
auto det_model_file = det_model_dir + sep + "inference.pdmodel";
|
||||
auto det_params_file = det_model_dir + sep + "inference.pdiparams";
|
||||
|
||||
auto cls_model_file = cls_model_dir + sep + "inference.pdmodel";
|
||||
auto cls_params_file = cls_model_dir + sep + "inference.pdiparams";
|
||||
|
||||
auto rec_model_file = rec_model_dir + sep + "inference.pdmodel";
|
||||
auto rec_params_file = rec_model_dir + sep + "inference.pdiparams";
|
||||
auto rec_label = rec_label_file;
|
||||
|
||||
fastdeploy::vision::ocr::DBDetector det_model;
|
||||
fastdeploy::vision::ocr::Classifier cls_model;
|
||||
fastdeploy::vision::ocr::Recognizer rec_model;
|
||||
|
||||
if (!det_model_dir.empty()) {
|
||||
auto det_option = fastdeploy::RuntimeOption();
|
||||
det_option.UseCpu();
|
||||
det_model = fastdeploy::vision::ocr::DBDetector(
|
||||
det_model_file, det_params_file, det_option);
|
||||
|
||||
if (!det_model.Initialized()) {
|
||||
std::cerr << "Failed to initialize det_model." << std::endl;
|
||||
return;
|
||||
}
|
||||
}
|
||||
|
||||
if (!cls_model_dir.empty()) {
|
||||
auto cls_option = fastdeploy::RuntimeOption();
|
||||
cls_option.UseCpu();
|
||||
cls_model = fastdeploy::vision::ocr::Classifier(
|
||||
cls_model_file, cls_params_file, cls_option);
|
||||
|
||||
if (!cls_model.Initialized()) {
|
||||
std::cerr << "Failed to initialize cls_model." << std::endl;
|
||||
return;
|
||||
}
|
||||
}
|
||||
|
||||
if (!rec_model_dir.empty()) {
|
||||
auto rec_option = fastdeploy::RuntimeOption();
|
||||
rec_option.UseCpu();
|
||||
rec_option.UsePaddleBackend(); // OCRv2的rec模型暂不支持ORT后端
|
||||
|
||||
rec_model = fastdeploy::vision::ocr::Recognizer(
|
||||
rec_model_file, rec_params_file, rec_label, rec_option);
|
||||
|
||||
if (!rec_model.Initialized()) {
|
||||
std::cerr << "Failed to initialize rec_model." << std::endl;
|
||||
return;
|
||||
}
|
||||
}
|
||||
|
||||
auto ocrv2_app = fastdeploy::application::ocrsystem::PPOCRSystemv2(
|
||||
&det_model, &cls_model, &rec_model);
|
||||
|
||||
auto im = cv::imread(image_file);
|
||||
auto im_bak = im.clone();
|
||||
|
||||
fastdeploy::vision::OCRResult res;
|
||||
//开始预测
|
||||
if (!ocrv2_app.Predict(&im, &res)) {
|
||||
std::cerr << "Failed to predict." << std::endl;
|
||||
return;
|
||||
}
|
||||
|
||||
//输出预测信息
|
||||
std::cout << res.Str() << std::endl;
|
||||
|
||||
//可视化
|
||||
auto vis_img = fastdeploy::vision::Visualize::VisOcr(im_bak, res);
|
||||
|
||||
cv::imwrite("vis_result.jpg", vis_img);
|
||||
std::cout << "Visualized result saved in ./vis_result.jpg" << std::endl;
|
||||
}
|
||||
|
||||
void GpuInfer(const std::string& det_model_dir,
|
||||
const std::string& cls_model_dir,
|
||||
const std::string& rec_model_dir,
|
||||
const std::string& rec_label_file,
|
||||
const std::string& image_file) {
|
||||
auto det_model_file = det_model_dir + sep + "inference.pdmodel";
|
||||
auto det_params_file = det_model_dir + sep + "inference.pdiparams";
|
||||
|
||||
auto cls_model_file = cls_model_dir + sep + "inference.pdmodel";
|
||||
auto cls_params_file = cls_model_dir + sep + "inference.pdiparams";
|
||||
|
||||
auto rec_model_file = rec_model_dir + sep + "inference.pdmodel";
|
||||
auto rec_params_file = rec_model_dir + sep + "inference.pdiparams";
|
||||
auto rec_label = rec_label_file;
|
||||
|
||||
fastdeploy::vision::ocr::DBDetector det_model;
|
||||
fastdeploy::vision::ocr::Classifier cls_model;
|
||||
fastdeploy::vision::ocr::Recognizer rec_model;
|
||||
|
||||
//准备模型
|
||||
if (!det_model_dir.empty()) {
|
||||
auto det_option = fastdeploy::RuntimeOption();
|
||||
det_option.UseGpu();
|
||||
det_model = fastdeploy::vision::ocr::DBDetector(
|
||||
det_model_file, det_params_file, det_option);
|
||||
|
||||
if (!det_model.Initialized()) {
|
||||
std::cerr << "Failed to initialize det_model." << std::endl;
|
||||
return;
|
||||
}
|
||||
}
|
||||
|
||||
if (!cls_model_dir.empty()) {
|
||||
auto cls_option = fastdeploy::RuntimeOption();
|
||||
cls_option.UseGpu();
|
||||
cls_model = fastdeploy::vision::ocr::Classifier(
|
||||
cls_model_file, cls_params_file, cls_option);
|
||||
|
||||
if (!cls_model.Initialized()) {
|
||||
std::cerr << "Failed to initialize cls_model." << std::endl;
|
||||
return;
|
||||
}
|
||||
}
|
||||
|
||||
if (!rec_model_dir.empty()) {
|
||||
auto rec_option = fastdeploy::RuntimeOption();
|
||||
rec_option.UseGpu();
|
||||
rec_option
|
||||
.UsePaddleBackend(); // OCRv2的rec模型暂不支持ORT后端与PaddleInference
|
||||
// v2.3.2
|
||||
rec_model = fastdeploy::vision::ocr::Recognizer(
|
||||
rec_model_file, rec_params_file, rec_label, rec_option);
|
||||
|
||||
if (!rec_model.Initialized()) {
|
||||
std::cerr << "Failed to initialize rec_model." << std::endl;
|
||||
return;
|
||||
}
|
||||
}
|
||||
|
||||
auto ocrv2_app = fastdeploy::application::ocrsystem::PPOCRSystemv2(
|
||||
&det_model, &cls_model, &rec_model);
|
||||
|
||||
auto im = cv::imread(image_file);
|
||||
auto im_bak = im.clone();
|
||||
|
||||
fastdeploy::vision::OCRResult res;
|
||||
//开始预测
|
||||
if (!ocrv2_app.Predict(&im, &res)) {
|
||||
std::cerr << "Failed to predict." << std::endl;
|
||||
return;
|
||||
}
|
||||
//输出预测信息
|
||||
std::cout << res.Str() << std::endl;
|
||||
|
||||
//可视化
|
||||
auto vis_img = fastdeploy::vision::Visualize::VisOcr(im_bak, res);
|
||||
|
||||
cv::imwrite("vis_result.jpg", vis_img);
|
||||
std::cout << "Visualized result saved in ./vis_result.jpg" << std::endl;
|
||||
}
|
||||
|
||||
void TrtInfer(const std::string& det_model_dir,
|
||||
const std::string& cls_model_dir,
|
||||
const std::string& rec_model_dir,
|
||||
const std::string& rec_label_file,
|
||||
const std::string& image_file) {
|
||||
auto det_model_file = det_model_dir + sep + "inference.pdmodel";
|
||||
auto det_params_file = det_model_dir + sep + "inference.pdiparams";
|
||||
|
||||
auto cls_model_file = cls_model_dir + sep + "inference.pdmodel";
|
||||
auto cls_params_file = cls_model_dir + sep + "inference.pdiparams";
|
||||
|
||||
auto rec_model_file = rec_model_dir + sep + "inference.pdmodel";
|
||||
auto rec_params_file = rec_model_dir + sep + "inference.pdiparams";
|
||||
auto rec_label = rec_label_file;
|
||||
|
||||
fastdeploy::vision::ocr::DBDetector det_model;
|
||||
fastdeploy::vision::ocr::Classifier cls_model;
|
||||
fastdeploy::vision::ocr::Recognizer rec_model;
|
||||
|
||||
//准备模型
|
||||
if (!det_model_dir.empty()) {
|
||||
auto det_option = fastdeploy::RuntimeOption();
|
||||
det_option.UseGpu();
|
||||
det_option.UseTrtBackend();
|
||||
det_option.SetTrtInputShape("x", {1, 3, 50, 50}, {1, 3, 640, 640},
|
||||
{1, 3, 960, 960});
|
||||
|
||||
det_model = fastdeploy::vision::ocr::DBDetector(
|
||||
det_model_file, det_params_file, det_option);
|
||||
|
||||
if (!det_model.Initialized()) {
|
||||
std::cerr << "Failed to initialize det_model." << std::endl;
|
||||
return;
|
||||
}
|
||||
}
|
||||
|
||||
if (!cls_model_dir.empty()) {
|
||||
auto cls_option = fastdeploy::RuntimeOption();
|
||||
cls_option.UseGpu();
|
||||
cls_option.UseTrtBackend();
|
||||
cls_option.SetTrtInputShape("x", {1, 3, 48, 192});
|
||||
|
||||
cls_model = fastdeploy::vision::ocr::Classifier(
|
||||
cls_model_file, cls_params_file, cls_option);
|
||||
|
||||
if (!cls_model.Initialized()) {
|
||||
std::cerr << "Failed to initialize cls_model." << std::endl;
|
||||
return;
|
||||
}
|
||||
}
|
||||
|
||||
if (!rec_model_dir.empty()) {
|
||||
auto rec_option = fastdeploy::RuntimeOption();
|
||||
rec_option.UseGpu();
|
||||
rec_option.UseTrtBackend();
|
||||
rec_option.SetTrtInputShape("x", {1, 3, 48, 10}, {1, 3, 48, 320},
|
||||
{1, 3, 48, 2000});
|
||||
|
||||
rec_model = fastdeploy::vision::ocr::Recognizer(
|
||||
rec_model_file, rec_params_file, rec_label, rec_option);
|
||||
|
||||
if (!rec_model.Initialized()) {
|
||||
std::cerr << "Failed to initialize rec_model." << std::endl;
|
||||
return;
|
||||
}
|
||||
}
|
||||
|
||||
auto ocrv2_app = fastdeploy::application::ocrsystem::PPOCRSystemv2(
|
||||
&det_model, &cls_model, &rec_model);
|
||||
|
||||
auto im = cv::imread(image_file);
|
||||
auto im_bak = im.clone();
|
||||
|
||||
fastdeploy::vision::OCRResult res;
|
||||
//开始预测
|
||||
if (!ocrv2_app.Predict(&im, &res)) {
|
||||
std::cerr << "Failed to predict." << std::endl;
|
||||
return;
|
||||
}
|
||||
//输出预测信息
|
||||
std::cout << res.Str() << std::endl;
|
||||
|
||||
//可视化
|
||||
auto vis_img = fastdeploy::vision::Visualize::VisOcr(im_bak, res);
|
||||
|
||||
cv::imwrite("vis_result.jpg", vis_img);
|
||||
std::cout << "Visualized result saved in ./vis_result.jpg" << std::endl;
|
||||
}
|
||||
|
||||
int main(int argc, char* argv[]) {
|
||||
if (argc < 7) {
|
||||
std::cout << "Usage: infer_demo path/to/det_model path/to/cls_model "
|
||||
"path/to/rec_model path/to/rec_label_file path/to/image "
|
||||
"run_option, "
|
||||
"e.g ./infer_demo ./ch_PP-OCRv2_det_infer "
|
||||
"./ch_ppocr_mobile_v2.0_cls_infer ./ch_PP-OCRv2_rec_infer "
|
||||
"./ppocr_keys_v1.txt ./12.jpg 0"
|
||||
<< std::endl;
|
||||
std::cout << "The data type of run_option is int, 0: run with cpu; 1: run "
|
||||
"with gpu; 2: run with gpu and use tensorrt backend."
|
||||
<< std::endl;
|
||||
return -1;
|
||||
}
|
||||
|
||||
if (std::atoi(argv[6]) == 0) {
|
||||
CpuInfer(argv[1], argv[2], argv[3], argv[4], argv[5]);
|
||||
} else if (std::atoi(argv[6]) == 1) {
|
||||
GpuInfer(argv[1], argv[2], argv[3], argv[4], argv[5]);
|
||||
} else if (std::atoi(argv[6]) == 2) {
|
||||
TrtInfer(argv[1], argv[2], argv[3], argv[4], argv[5]);
|
||||
}
|
||||
return 0;
|
||||
}
|
||||
Reference in New Issue
Block a user