Refine code structure (#89)

* refine code structure * refine code structure
2025-10-13 12:23:55 +08:00 · 2022-08-10 10:50:22 +08:00
parent c7d37b6732
commit 22ca63982b
333 changed files with 1 additions and 37500 deletions
--- a/examples/vision/detection/README.md
+++ b/examples/vision/detection/README.md
@@ -0,0 +1,23 @@
+# 视觉模型部署
+
+本目录下提供了各类视觉模型的部署，主要涵盖以下任务类型
+
+| 任务类型           | 说明                                  | 预测结果结构体                                                                          |
+|:-------------- |:----------------------------------- |:-------------------------------------------------------------------------------- |
+| Detection      | 目标检测，输入图像，检测图像中物体位置，并返回检测框坐标及类别和置信度 | [DetectionResult](../../../../docs/api/vision_results/detection_result.md)       |
+| Segmentation   | 语义分割，输入图像，给出图像中每个像素的分类及置信度          | [SegmentationResult](../../../../docs/api/vision_results/segmentation_result.md) |
+| Classification | 图像分类，输入图像，给出图像的分类结果和置信度             | [ClassifyResult](../../../../docs/api/vision_results/classification_result.md)   |
+
+## FastDeploy API设计
+
+视觉模型具有较有统一任务范式，在设计API时（包括C++/Python），FastDeploy将视觉模型的部署拆分为四个步骤
+
+- 模型加载
+- 图像预处理
+- 模型推理
+- 推理结果后处理
+
+FastDeploy针对飞桨的视觉套件，以及外部热门模型，提供端到端的部署服务，用户只需准备模型，按以下步骤即可完成整个模型的部署
+
+- 加载模型
+- 调用`predict`接口
--- a/examples/vision/detection/yolov7/README.md
+++ b/examples/vision/detection/yolov7/README.md
@@ -0,0 +1,39 @@
+# YOLOv7准备部署模型
+
+## 模型版本说明
+
+- [YOLOv7 0.1](https://github.com/WongKinYiu/yolov7/releases/tag/v0.1)
+  - （1）[链接中](https://github.com/WongKinYiu/yolov7/releases/tag/v0.1)的*.pt通过[导出ONNX模型](#导出ONNX模型)操作后，可进行部署；
+  - （2）[链接中](https://github.com/WongKinYiu/yolov7/releases/tag/v0.1)的*.onnx、*.trt和 *.pose模型不支持部署；
+  - （3）开发者基于自己数据训练的YOLOv7 0.1模型，可按照[导出ONNX模型](#%E5%AF%BC%E5%87%BAONNX%E6%A8%A1%E5%9E%8B)后，完成部署。
+
+## 导出ONNX模型
+
+```
+# 下载yolov7模型文件
+wget https://github.com/WongKinYiu/yolov7/releases/download/v0.1/yolov7.pt
+
+# 导出onnx格式文件 (Tips: 对应 YOLOv7 release v0.1 代码)
+python models/export.py --grid --dynamic --weights PATH/TO/yolov7.pt
+
+# 如果您的代码版本中有支持NMS的ONNX文件导出，请使用如下命令导出ONNX文件(请暂时不要使用 "--end2end"，我们后续将支持带有NMS的ONNX模型的部署)
+python models/export.py --grid --dynamic --weights PATH/TO/yolov7.pt
+
+# 移动onnx文件到demo目录
+cp PATH/TO/yolov7.onnx PATH/TO/model_zoo/vision/yolov7/
+```
+
+## 下载预训练模型
+
+为了方便开发者的测试，下面提供了YOLOv7导出的各系列模型，开发者可直接下载使用。
+
+| 模型                                                               | 大小    | 精度    |
+|:---------------------------------------------------------------- |:----- |:----- |
+| [YOLOv7](https://bj.bcebos.com/paddlehub/fastdeploy/yolov7.onnx) | 141MB | 51.4% |
+| [YOLOv7-x]                                                       | 10MB  | 51.4% |
+
+
+## 详细部署文档
+
+- [Python部署](python)
+- [C++部署](cpp)
--- a/examples/vision/detection/yolov7/cpp/CMakeLists.txt
+++ b/examples/vision/detection/yolov7/cpp/CMakeLists.txt
@@ -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 ${FASTDEPLOY_LIBS})
--- a/examples/vision/detection/yolov7/cpp/README.md
+++ b/examples/vision/detection/yolov7/cpp/README.md
@@ -0,0 +1,77 @@
+# YOLOv7 C++部署示例
+
+本目录下提供`infer.cc`快速完成YOLOv7在CPU/GPU，以及GPU上通过TensorRT加速部署的示例。
+
+在部署前，需确认以下两个步骤
+
+- 1. 软硬件环境满足要求，参考[FastDeploy环境要求](../../../../../docs/quick_start/requirements.md)  
+- 2. 根据开发环境，下载预编译部署库和samples代码，参考[FastDeploy预编译库](../../../../../docs/compile/prebuild_libraries.md)
+
+以Linux上CPU推理为例，在本目录执行如下命令即可完成编译测试
+
+```
+mkdir build
+cd build
+wget https://xxx.tgz
+tar xvf fastdeploy-linux-x64-0.2.0.tgz
+cmake .. -DFASTDEPLOY_INSTALL_DIR=${PWD}/fastdeploy-linux-x64-0.2.0
+make -j
+
+#下载官方转换好的yolov7模型文件和测试图片
+wget https://bj.bcebos.com/paddlehub/fastdeploy/yolov7.onnx
+wget https://gitee.com/paddlepaddle/PaddleDetection/raw/release/2.4/demo/000000087038.jpg
+
+
+# CPU推理
+./infer_demo yolov7.onnx 000000087038.jpg 0
+# GPU推理
+./infer_demo yolov7.onnx 000000087038.jpg 1
+# GPU上TensorRT推理
+./infer_demo yolov7.onnx 000000087038.jpg 2
+```
+
+## YOLOv7 C++接口
+
+### YOLOv7类
+
+```
+fastdeploy::vision::detection::YOLOv7(
+        const string& model_file,
+        const string& params_file = "",
+        const RuntimeOption& runtime_option = RuntimeOption(),
+        const Frontend& model_format = Frontend::ONNX)
+```
+
+YOLOv7模型加载和初始化，其中model_file为导出的ONNX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ONNX格式
+
+#### Predict函数
+
+> ```
+> YOLOv7::Predict(cv::Mat* im, DetectionResult* result,
+>                 float conf_threshold = 0.25,
+>                 float nms_iou_threshold = 0.5)
+> ```
+> 
+> 模型预测接口，输入图像直接输出检测结果。
+> 
+> **参数**
+> 
+> > * **im**: 输入图像，注意需为HWC，BGR格式
+> > * **result**: 检测结果，包括检测框，各个框的置信度, DetectionResult说明参考[视觉模型预测结果](../../../../../docs/api/vision_results/)
+> > * **conf_threshold**: 检测框置信度过滤阈值
+> > * **nms_iou_threshold**: NMS处理过程中iou阈值
+
+### 类成员变量
+
+> > * **size**(vector<int>): 通过此参数修改预处理过程中resize的大小，包含两个整型元素，表示[width, height], 默认值为[640, 640]
+
+- [模型介绍](../../)
+- [Python部署](../python)
+- [视觉模型预测结果](../../../../../docs/api/vision_results/)
--- a/examples/vision/detection/yolov7/cpp/infer.cc
+++ b/examples/vision/detection/yolov7/cpp/infer.cc
@@ -0,0 +1,105 @@
+// 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"
+
+void CpuInfer(const std::string& model_file, const std::string& image_file) {
+  auto model = fastdeploy::vision::detection::YOLOv7(model_file);
+  if (!model.Initialized()) {
+    std::cerr << "Failed to initialize." << std::endl;
+    return;
+  }
+
+  auto im = cv::imread(image_file);
+  auto im_bak = im.clone();
+
+  fastdeploy::vision::DetectionResult res;
+  if (!model.Predict(&im, &res)) {
+    std::cerr << "Failed to predict." << std::endl;
+    return;
+  }
+
+  auto vis_im = fastdeploy::vision::Visualize::VisDetection(im_bak, res);
+  cv::imwrite("vis_result.jpg", vis_im);
+  std::cout << "Visualized result saved in ./vis_result.jpg" << std::endl;
+}
+
+void GpuInfer(const std::string& model_file, const std::string& image_file) {
+  auto option = fastdeploy::RuntimeOption();
+  option.UseGpu();
+  auto model = fastdeploy::vision::detection::YOLOv7(model_file, "", option);
+  if (!model.Initialized()) {
+    std::cerr << "Failed to initialize." << std::endl;
+    return;
+  }
+
+  auto im = cv::imread(image_file);
+  auto im_bak = im.clone();
+
+  fastdeploy::vision::DetectionResult res;
+  if (!model.Predict(&im, &res)) {
+    std::cerr << "Failed to predict." << std::endl;
+    return;
+  }
+
+  auto vis_im = fastdeploy::vision::Visualize::VisDetection(im_bak, res);
+  cv::imwrite("vis_result.jpg", vis_im);
+  std::cout << "Visualized result saved in ./vis_result.jpg" << std::endl;
+}
+
+void TrtInfer(const std::string& model_file, const std::string& image_file) {
+  auto option = fastdeploy::RuntimeOption();
+  option.UseGpu();
+  option.UseTrtBackend();
+  option.SetTrtInputShape("images", {1, 3, 640, 640});
+  auto model = fastdeploy::vision::detection::YOLOv7(model_file, "", option);
+  if (!model.Initialized()) {
+    std::cerr << "Failed to initialize." << std::endl;
+    return;
+  }
+
+  auto im = cv::imread(image_file);
+  auto im_bak = im.clone();
+
+  fastdeploy::vision::DetectionResult res;
+  if (!model.Predict(&im, &res)) {
+    std::cerr << "Failed to predict." << std::endl;
+    return;
+  }
+
+  auto vis_im = fastdeploy::vision::Visualize::VisDetection(im_bak, res);
+  cv::imwrite("vis_result.jpg", vis_im);
+  std::cout << "Visualized result saved in ./vis_result.jpg" << std::endl;
+}
+
+int main(int argc, char* argv[]) {
+  if (argc < 4) {
+    std::cout << "Usage: infer_demo path/to/model path/to/image run_option, "
+                 "e.g ./infer_model ./yolov7.onnx ./test.jpeg 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[3]) == 0) {
+    CpuInfer(argv[1], argv[2]);
+  } else if (std::atoi(argv[3]) == 1) {
+    GpuInfer(argv[1], argv[2]);
+  } else if (std::atoi(argv[3]) == 2) {
+    TrtInfer(argv[1], argv[2]);
+  }
+  return 0;
+}
--- a/examples/vision/detection/yolov7/python/README.md
+++ b/examples/vision/detection/yolov7/python/README.md
@@ -0,0 +1,71 @@
+# YOLOv7 Python部署示例
+
+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/quick_start/requirements.md)  
+- 2. FastDeploy Python whl包安装，参考[FastDeploy Python安装](../../../../../docs/quick_start/install.md)
+
+本目录下提供`infer.py`快速完成YOLOv7在CPU/GPU，以及GPU上通过TensorRT加速部署的示例。执行如下脚本即可完成
+
+```
+#下载yolov7模型文件和测试图片
+wget https://bj.bcebos.com/paddlehub/fastdeploy/yolov7.onnx
+wget https://gitee.com/paddlepaddle/PaddleDetection/raw/release/2.4/demo/000000014439.jpg
+
+
+#下载部署示例代码
+git clone https://github.com/PaddlePaddle/FastDeploy.git
+cd examples/vison/detection/yolov7/python/
+
+# CPU推理
+python infer.py --model yolov7.onnx --image 000000087038.jpg --device cpu
+# GPU推理
+python infer.py --model yolov7.onnx --image 000000087038.jpg --device gpu
+# GPU上使用TensorRT推理
+python infer.py --model yolov7.onnx --image 000000087038.jpg --device gpu --use_trt True
+```
+
+运行完成可视化结果如下图所示
+
+## YOLOv7 Python接口
+
+```
+fastdeploy.vision.detection.YOLOv7(model_file, params_file=None, runtime_option=None, model_format=Frontend.ONNX)
+```
+
+YOLOv7模型加载和初始化，其中model_file为导出的ONNX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格式时，此参数无需设定
+> * **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ONNX
+
+### predict函数
+
+> ```
+> YOLOv7.predict(image_data, conf_threshold=0.25, nms_iou_threshold=0.5)
+> ```
+> 
+> 模型预测结口，输入图像直接输出检测结果。
+> 
+> **参数**
+> 
+> > * **image_data**(np.ndarray): 输入数据，注意需为HWC，BGR格式
+> > * **conf_threshold**(float): 检测框置信度过滤阈值
+> > * **nms_iou_threshold**(float): NMS处理过程中iou阈值
+
+> **返回**
+> 
+> > 返回`fastdeploy.vision.DetectionResult`结构体，结构体说明参考文档[视觉模型预测结果](../../../../../docs/api/vision_results/)
+
+### 类成员属性
+
+> > * **size**(list | tuple): 通过此参数修改预处理过程中resize的大小，包含两个整型元素，表示[width, height], 默认值为[640, 640]
+
+## 其它文档
+
+- [YOLOv7 模型介绍](..)
+- [YOLOv7 C++部署](../cpp)
+- [模型预测结果说明](../../../../../docs/api/vision_results/)
--- a/examples/vision/detection/yolov7/python/infer.py
+++ b/examples/vision/detection/yolov7/python/infer.py
@@ -0,0 +1,51 @@
+import fastdeploy as fd
+import cv2
+
+
+def parse_arguments():
+    import argparse
+    import ast
+    parser = argparse.ArgumentParser()
+    parser.add_argument(
+        "--model", required=True, help="Path of yolov7 onnx model.")
+    parser.add_argument(
+        "--image", required=True, help="Path of test image file.")
+    parser.add_argument(
+        "--device",
+        type=str,
+        default='cpu',
+        help="Type of inference device, support 'cpu' or 'gpu'.")
+    parser.add_argument(
+        "--use_trt",
+        type=ast.literal_eval,
+        default=False,
+        help="Wether to use tensorrt.")
+    return parser.parse_args()
+
+
+def build_option(args):
+    option = fd.RuntimeOption()
+
+    if args.device.lower() == "gpu":
+        option.use_gpu()
+
+    if args.use_trt:
+        option.use_trt_backend()
+        option.set_trt_input_shape("images", [1, 3, 640, 640])
+    return option
+
+
+args = parse_arguments()
+
+# 配置runtime，加载模型
+runtime_option = build_option(args)
+model = fd.vision.detection.YOLOv7(args.model, runtime_option=runtime_option)
+
+# 预测图片检测结果
+im = cv2.imread(args.image)
+result = model.predict(im)
+
+# 预测结果可视化
+vis_im = fd.vision.vis_detection(im, result)
+cv2.imwrite("visualized_result.jpg", vis_im)
+print("Visualized result save in ./visualized_result.jpg")