From 88d7ff971b792bdef6169ef82794eba36d4cad41 Mon Sep 17 00:00:00 2001 From: BaofengZan Date: Fri, 28 Aug 2020 13:17:39 +0800 Subject: [PATCH] =?UTF-8?q?windows=E7=BC=96=E8=AF=91=E8=BF=90=E8=A1=8C?= =?UTF-8?q?=E6=95=99=E7=A8=8B?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit --- windows编译运行手把手教程.md | 150 +++++++++++++++++++++++++++++++++++ 1 file changed, 150 insertions(+) create mode 100644 windows编译运行手把手教程.md diff --git a/windows编译运行手把手教程.md b/windows编译运行手把手教程.md new file mode 100644 index 0000000..05678b7 --- /dev/null +++ b/windows编译运行手把手教程.md @@ -0,0 +1,150 @@ +# windows编译运行手把手教程 + +## 说明 + +本教程可以用于https://github.com/wang-xinyu/tensorrtx该仓库中所有项目，只需要修改几处代码即可，修改之处后面会详细注明。 + +## 所需环境 + +* vs 版本可不限（仅测试了 vs2015 vs2017） + +* cuda + +* TensorRT需要和本机的cuda版本适配（需要7.0以上版本） + +* Cmake + +* opencv （需要和vs版本适配） + +* 增加windows下的文件夹处理头文件dirent.h，放在当前文件下include下。下载地址 https://github.com/tronkko/dirent + + ![image-20200828131208257](imgs/image-20200828131208257.png) + +## 编译 + +### 修改CmakeLists.txt + +```cmake +cmake_minimum_required(VERSION 2.6) + +project(yolov5) # 1 +set(OpenCV_DIR "D:\\opencv\\opencv346\\build") #2 +set(TRT_DIR "D:\\TensorRT-7.0.0.11.Windows10.x86_64.cuda-10.2.cudnn7.6\\TensorRT-7.0.0.11") #3 + +add_definitions(-std=c++11) +option(CUDA_USE_STATIC_CUDA_RUNTIME OFF) +set(CMAKE_CXX_STANDARD 11) +set(CMAKE_BUILD_TYPE Debug) + +set(THREADS_PREFER_PTHREAD_FLAG ON) +find_package(Threads) + +#设置cuda信息 +find_package(CUDA REQUIRED) +message(STATUS " libraries: ${CUDA_LIBRARIES}") +message(STATUS " include path: ${CUDA_INCLUDE_DIRS}") + +include_directories(${CUDA_INCLUDE_DIRS}) + +set(CUDA_NVCC_PLAGS ${CUDA_NVCC_PLAGS};-std=c++11; -g; -G;-gencode; arch=compute_75;code=sm_75) +#### +enable_language(CUDA) # 这一句添加后 ，就会在vs中不需要再手动设置cuda +#### +include_directories(${PROJECT_SOURCE_DIR}/include) +include_directories(${TRT_DIR}\\include) + +#-D_MWAITXINTRIN_H_INCLUDED 解决error: identifier "__builtin_ia32_mwaitx" is undefined +set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -std=c++11 -Wall -Ofast -D_MWAITXINTRIN_H_INCLUDED") + +# 设置opencv的信息 +find_package(OpenCV QUIET + NO_MODULE + NO_DEFAULT_PATH + NO_CMAKE_PATH + NO_CMAKE_ENVIRONMENT_PATH + NO_SYSTEM_ENVIRONMENT_PATH + NO_CMAKE_PACKAGE_REGISTRY + NO_CMAKE_BUILDS_PATH + NO_CMAKE_SYSTEM_PATH + NO_CMAKE_SYSTEM_PACKAGE_REGISTRY +) + +message(STATUS "OpenCV library status:") +message(STATUS " version: ${OpenCV_VERSION}") +message(STATUS " libraries: ${OpenCV_LIBS}") +message(STATUS " include path: ${OpenCV_INCLUDE_DIRS}") + +include_directories(${OpenCV_INCLUDE_DIRS}) +link_directories(${TRT_DIR}\\lib) + +add_executable(yolov5 ${PROJECT_SOURCE_DIR}/yolov5.cpp ${PROJECT_SOURCE_DIR}/yololayer.cu ${PROJECT_SOURCE_DIR}/yololayer.h + ${PROJECT_SOURCE_DIR}/hardswish.cu ${PROJECT_SOURCE_DIR}/hardswish.h) #4 + +target_link_libraries(yolov5 "nvinfer" "nvinfer_plugin") #5 +target_link_libraries(yolov5 ${OpenCV_LIBS}) #6 +target_link_libraries(yolov5 ${CUDA_LIBRARIES}) #7 +target_link_libraries(yolov5 Threads::Threads) #8 +``` + +注意：在CamakeLists.txt中，有8处需要修改，在#1-#8标注的位置 + +1. #1 为当前编译的工程名，根据你所编译的工程名修改，也可以自定义 +2. #2 为本机的opencv地址 +3. #3 为本机TensorRT的安装位置 +4. #4 为当前编译工程中所涉及的文件，包括.cpp .cu .h等文件，另外第一个参数要与#1设置的相同 +5. #5 -#8 为链接lib，工程名需要和#1设置的一样。 + +### cmake-gui编译 + +#### 1 打开cmake-gui，并设置相应路径：如下图 + +![image-20200828124434245](imgs/image-20200828124434245.png) + +#### 点击1处**Configure**，弹出窗口，并选择相应环境： + +![image-20200828124902923](imgs/image-20200828124902923.png) + +点击Finish完成设置，等待生成完成： + +![image-20200828124951872](imgs/image-20200828124951872.png) + +#### 点击2处Generate + +![image-20200828125046738](imgs/image-20200828125046738.png) + +#### 点击3处Open Project，打开工程 + +![image-20200828125215067](imgs/image-20200828125215067.png) + +## 运行 + +点击“生成-生成解决方案”。等待编译完成。 + +![image-20200828125402056](imgs/image-20200828125402056.png) + +### 运行方法1：命令行 + +命令行cd到exe的路径（比如：E:\LearningCodes\GithubRepo\tensorrtx\yolov5\build\Debug）, + +``` +yolov5.exe -s // serialize model to plan file i.e. 'yolov5s.engine' +yolov5.exe -d ../samples // deserialize plan file and run inference, the images in samples will be processed. +``` + +**注意：在生成engine时，wts文件要放置在工程文件xxx.vcxproj所在文件的上一级目录**，或者直接修改源码，直接指向绝对地址： + +![image-20200828125938472](imgs/image-20200828125938472.png) + +### 运行方法2：vs直接运行，可debug + +在vs中，在工程上右键，先选择“设为启动项”，然后选择“属性”。在属性页设置参数。设置完成后点击运行即可，可以打断点debug。 + +![image-20200828130117902](imgs/image-20200828130117902.png) + +![image-20200828130415658](imgs/image-20200828130415658.png) + +![image-20200828131516231](imgs/image-20200828131516231.png) + +**注意：** + +**运行时需要将tensorRt的dll和opencv的dll拷贝到exe所在路径。或者将tensorRt和opencv的bin文件夹路径添加到环境变量中（不建议）** \ No newline at end of file