windows：Opencv使用Cmake & VS 编译， C++、python、Cuda、Qt 环境详细教学及踩坑一：编译篇

Bethxone

已于 2024-07-19 18:38:18 修改

阅读量689

点赞数 24

分类专栏： Opencv cmake 文章标签： opencv c++ python

于 2024-07-19 16:16:34 首次发布

本文链接：https://blog.csdn.net/Bethxone/article/details/140526614

版权

Opencv cmake 专栏收录该内容

6 篇文章 0 订阅

订阅专栏

一、准备：

本文使用的软件版本均是撰写时的最新或次新版，对过旧版本不一定通用

0、环境配置：

Windows	Windows 11 专业版 23H2
opencv	4.10
Qt	6.7.2
Cmake	3.3.0
python	3.12.4
cuda	12.5
cudnn	9.2.1
vs studio	2022

1、 Opencv，Opencv_contrib 源码下载本文版本：4.10

GitHub - opencv/opencv: Open Source Computer Vision Library

GitHub - opencv/opencv_contrib: Repository for OpenCV's extra modules

下载完成后解压opencv 和 opencv_contrib，将opencv_contrib文件夹移入opencv文件夹下，并在opencv文件夹下新建build文件夹。

软件安装教程引用其他大神的教程，很详细

1、选择自定义安装

2、选择添加python路径到环境变量

3、选择为所有用户安装

其他选项如图所示，如果cmake需要debug版本的一定要勾选debug有关的选项，否则生成过程中会缺少python3xx_d.lib 文件。

5、Cuda 、cudNN 安装：本文版本：Cuda 12.5 cudNN 9.2

先按该教程安装cuda、cudnn：Cuda和cuDNN安装教程(超级详细)-CSDN博客

cudnn再按官方文档安装一遍：英伟达官方cudnn安装教程

注意：cudnn一定要按教程，将cudnn文件复制到cuda文件夹下。

否则：如果只是按照官方文档中的教程，编译成功后，python import cv2 时无法调用cudnn中的dll文件。报错：

ImportError: DLL load failed: The specified module could not be found.

！！注意：Windows下使用cmake编译cuda只能使用VS，否则会无法找到cuda位置

6、VS studio 安装：

Thank You for Downloading Visual Studio Community Edition

7、上述软件安装完成后重启计算机！！！

二、设置 Cmake 选项

参考：AMD Customer Community

从“开始”菜单打开 CMake （cmake-gui）
将“Where is the source code”设置为提取的 opencv 文件夹。
将“Where to build the binaries”设置为提取的 opencv 文件夹，并在末尾附加 /build。这将是编译文件的构建位置，您可以将其设置为所需的任何文件夹
点击“configure”
弹出窗口选择vs2019作为编译器
generate平台选择x64

注：此处 opencv_cp312_410、opencv_410_cp312_amd64_bulid、opencv_410_cp312_amd64为笔者按习惯重命名。

对应关系为：

                opencv_cp312_410 ——>opencv_x.xx opencv主文件夹

                opencv_410_cp312_amd64_bulid ——> build

opencv_410_cp312_amd64  ——> install

此处的目录结构为：

----opencv_x.xx

----3rdparty

---- .....

----build

----install

---- .....

---- opencv_contrib

1、通用编译选项：

参考链接：AMD Customer Community

使用“搜索”框更改以下属性：

BUILD_PROTOBUF：– 取消选中
WITH_PROTOBUF：– 取消选中
BUILD_TESTS：– 取消选中
WITH_OPENEXR：– 取消选中
BUILD_OPENEXR：– 取消选中
OPENCV_ENABLE_ALLOCATOR_STATS：– 取消选中
CMAKE_BUILD_TYPE：– RELEASE
CMAKE_INSTALL_PREFIX：– 这默认为 build/install 目录，但您可以将其更改为您希望将 OpenCV 库安装到的任何目录。
BUILD_opencv_world：建议勾选，build_opencv_world可以将所有opencv的库都打包编译在一起。
OPENCV_ENABLE_NONFREE：勾选，否则无法使用sift。
WITH_OPENGL ：勾选
OPENCV_EXTRA_MODULES_PATH – 设置为提取的 opencv_contrib/modules 文件夹的路径。如图所示：

10、配置完成点击“configure”

报错信息1：

CMake Warning (dev) at CMakeLists.txt:127 (enable_language):
project() should be called prior to this enable_language() call.
This warning is for project developers. Use -Wno-dev to suppress it.

Selecting Windows SDK version 10.0.22621.0 to target Windows 10.0.22631.
CMake Warning (dev) at C:/Program Files/CMake/share/cmake-3.30/Modules/Platform/Windows-MSVC.cmake:539 (enable_language):
project() should be called prior to this enable_language() call.
Call Stack (most recent call first):
C:/Program Files/CMake/share/cmake-3.30/Modules/Platform/Windows-MSVC.cmake:509 (__windows_compiler_msvc_enable_rc)
C:/Program Files/CMake/share/cmake-3.30/Modules/Platform/Windows-MSVC-CXX.cmake:6 (__windows_compiler_msvc)
C:/Program Files/CMake/share/cmake-3.30/Modules/CMakeCXXInformation.cmake:48 (include)
CMakeLists.txt:127 (enable_language)
This warning is for project developers. Use -Wno-dev to suppress it.

CMake Warning (dev) at C:/Program Files/CMake/share/cmake-3.30/Modules/Platform/Windows-MSVC.cmake:539 (enable_language):
project() should be called prior to this enable_language() call.
Call Stack (most recent call first):
C:/Program Files/CMake/share/cmake-3.30/Modules/Platform/Windows-MSVC.cmake:509 (__windows_compiler_msvc_enable_rc)
C:/Program Files/CMake/share/cmake-3.30/Modules/Platform/Windows-MSVC-C.cmake:5 (__windows_compiler_msvc)
C:/Program Files/CMake/share/cmake-3.30/Modules/CMakeCInformation.cmake:48 (include)
CMakeLists.txt:127 (enable_language)
This warning is for project developers. Use -Wno-dev to suppress it.

报错信息2：

CMake Warning at cmake/OpenCVGenSetupVars.cmake:54 (message):
CONFIGURATION IS NOT SUPPORTED: validate setupvars script in install
directory
Call Stack (most recent call first):
CMakeLists.txt:1180 (include)

解决方案：

这两类警告，不影响正常编译过程。对此未作处理。

2、cuda 编译选项

参考链接：windows11编译OpenCV4.5.0 with CUDA（附注意事项）-CSDN博客

第一次搜cuda时将带cuda的都勾选，点击“configure”

搜索：WITH_NVCUVENC——取消选中

WITH_NVCUVID——取消选中

注意：第一次搜索时，如图所示的部分选项不存在是正常现象，因为每次"configure"时，cmake都会根据已经勾选的选项重新加载，所以及时的重新"configure"是一个好习惯。

将已经存在的选项按图勾选后，再次点击"configure"后重新搜索，即可出现全部选项。

配置CUDA_ARCH_BIN中将显卡的算力内容改成自己显卡的算力,显卡算力查看链接查看自己显卡的算力，我的显卡是4060laptop，官网算力为8.9。CUDA_ARCH_BIN中将低于8.9数值的删除。

cuda编译选项按图进行勾选。

报错信息1：WITH_NVCUVENC、WITH_NVCUVID

CMake Warning at opencv_contrib_4_10/modules/cudacodec/CMakeLists.txt:26 (message):
cudacodec::VideoReader requires Nvidia Video Codec SDK. Please resolve
dependency or disable WITH_NVCUVID=OFF
Call Stack (most recent call first):
modules/world/CMakeLists.txt:13 (include)
modules/world/CMakeLists.txt:50 (include_one_module)
CMake Warning at opencv_contrib_4_10/modules/cudacodec/CMakeLists.txt:30 (message):
cudacodec::VideoWriter requires Nvidia Video Codec SDK. Please resolve
dependency or disable WITH_NVCUVENC=OFF
Call Stack (most recent call first):
modules/world/CMakeLists.txt:13 (include)
modules/world/CMakeLists.txt:50 (include_one_module)

解决方法：

搜索：WITH_NVCUVENC——取消选中

WITH_NVCUVID——取消选中

报错信息2：OPENCV_DNN_CUDA

CMake Error at modules/dnn/CMakeLists.txt:53 (message):
DNN: CUDA backend requires cuDNN. Please resolve dependency or disable
OPENCV_DNN_CUDA=OFF
Call Stack (most recent call first):
modules/world/CMakeLists.txt:13 (include)
modules/world/CMakeLists.txt:50 (include_one_module)

解决方法：

搜索CUDNN_LIBRARY，找到并输入cudnn.lib的安装位置

CUDA 参考配置：

打开 Advance，配置供参考（不一定要一致，只是提供参考）

3、python 编译选项

参考：【opencv】【GPU】windows10下opencv4.8.0-cuda Python版本源码编译教程_cv2的版本4.8.0-CSDN博客

注意：有的同学即使按图勾选后依然无法出现某些选项。

例如：

若无法出现BUILD_opencv_python3，（此选项为python编译的必需选项）此时可以进行手动添加。点击 Add Entry ，在Name 中输入BUILD_opencv_python3，点击OK，即可出现该选项。

python的选项设置如图所示，其中numpy需提前进行安装，安装完成python后，打开cmd，输入

pip install numpy

python 参考配置：

4、Qt 编译选项

参考：基于QT5.14.2的OPENCV4.8.0的cuda11.4开发环境搭建和源码编译_opencv4.8编译-CSDN博客

1、搜索Qt，设置Qt6_DIR目录为：Qt安装路径/6.7.2/msvc2019_64/lib/cmake/Qt6，如图所示。点击“configure”。

2、查找Qt安装路径\6.7.2\msvc2019_64\lib\cmake\Qtest文件夹，将其压缩备份后删除。

参考：https://github.com/opencv/opencv/issues/23826#issuecomment-2227167370

否则会报错。点击“configure”。 此时搜索 Qt6Test_DIR 为 Qt6Test_DIR-NOTFOUND。

报错信息：Qtest

CMake Error at C:/Program Files/CMake/share/cmake-3.30/Modules/FindCUDA.cmake:1793 (file):
Error evaluating generator expression:

$<TARGET_PROPERTY:QT_TESTCASE_BUILDDIR>

$<TARGET_PROPERTY:prop> may only be used with binary targets. It may not
be used with add_custom_command or add_custom_target. Specify the target
to read a property from using the $<TARGET_PROPERTY:tgt,prop> signature
instead.
Call Stack (most recent call first):
C:/Program Files/CMake/share/cmake-3.30/Modules/FindCUDA.cmake:2111 (CUDA_WRAP_SRCS)
C:/Program Files/CMake/share/cmake-3.30/Modules/FindCUDA.cmake:2124 (cuda_compile_base)
cmake/OpenCVDetectCUDA.cmake:164 (CUDA_COMPILE)
cmake/OpenCVUtils.cmake:1578 (ocv_cuda_compile)
cmake/OpenCVModule.cmake:989 (ocv_add_library)
cmake/OpenCVModule.cmake:905 (_ocv_create_module)
modules/world/CMakeLists.txt:78 (ocv_create_module)

解决方案：

查找Qt安装路径\6.7.2\msvc2019_64\lib\cmake\Qtest文件夹，将其压缩备份后删除。

参考：https://github.com/opencv/opencv/issues/23826#issuecomment-2227167370

Qt 参考配置：

注意：每次点击"configure"都要注意报错信息，除了一开始的两个警告外，其他警告一定要及时解决，否则编译会失败

三、检查配置报告

！！！检查具有背景色的行，看是否与自己预期一致。

General configuration for OpenCV 4.10.0 =====================================

Version control: unknown

Extra modules:

Location (extra): D:/Program Files/opencv/opencv_cp312_410/opencv_contrib_4_10/modules

Version control (extra): unknown

Platform:

Timestamp: 2024-07-19T05:23:16Z

Host: Windows 10.0.22631 AMD64

CMake: 3.30.0

CMake generator: Visual Studio 17 2022

CMake build tool: C:/Program Files (x86)/Microsoft Visual Studio/2022/BuildTools/MSBuild/Current/Bin/amd64/MSBuild.exe

MSVC: 1940

Configuration: Debug Release

CPU/HW features:

Baseline: SSE SSE2 SSE3

requested: SSE3

Dispatched code generation: SSE4_1 SSE4_2 FP16 AVX AVX2 AVX512_SKX

requested: SSE4_1 SSE4_2 AVX FP16 AVX2 AVX512_SKX

SSE4_1 (16 files): + SSSE3 SSE4_1

SSE4_2 (1 files): + SSSE3 SSE4_1 POPCNT SSE4_2

FP16 (0 files): + SSSE3 SSE4_1 POPCNT SSE4_2 FP16 AVX

AVX (8 files): + SSSE3 SSE4_1 POPCNT SSE4_2 AVX

AVX2 (36 files): + SSSE3 SSE4_1 POPCNT SSE4_2 FP16 FMA3 AVX AVX2

AVX512_SKX (5 files): + SSSE3 SSE4_1 POPCNT SSE4_2 FP16 FMA3 AVX AVX2 AVX_512F AVX512_COMMON AVX512_SKX

C/C++:

Built as dynamic libs?: YES

C++ standard: 11

C++ Compiler: C:/Program Files (x86)/Microsoft Visual Studio/2022/BuildTools/VC/Tools/MSVC/14.40.33807/bin/Hostx64/x64/cl.exe (ver 19.40.33812.0)

C++ flags (Release): /DWIN32 /D_WINDOWS /W4 /GR /D _CRT_SECURE_NO_DEPRECATE /D _CRT_NONSTDC_NO_DEPRECATE /D _SCL_SECURE_NO_WARNINGS /Gy /bigobj /Oi /fp:precise /EHa /wd4127 /wd4251 /wd4324 /wd4275 /wd4512 /wd4589 /wd4819 /MP /O2 /Ob2 /DNDEBUG

C++ flags (Debug): /DWIN32 /D_WINDOWS /W4 /GR /D _CRT_SECURE_NO_DEPRECATE /D _CRT_NONSTDC_NO_DEPRECATE /D _SCL_SECURE_NO_WARNINGS /Gy /bigobj /Oi /fp:precise /EHa /wd4127 /wd4251 /wd4324 /wd4275 /wd4512 /wd4589 /wd4819 /MP /Zi /Ob0 /Od /RTC1

C Compiler: C:/Program Files (x86)/Microsoft Visual Studio/2022/BuildTools/VC/Tools/MSVC/14.40.33807/bin/Hostx64/x64/cl.exe

C flags (Release): /DWIN32 /D_WINDOWS /W3 /D _CRT_SECURE_NO_DEPRECATE /D _CRT_NONSTDC_NO_DEPRECATE /D _SCL_SECURE_NO_WARNINGS /Gy /bigobj /Oi /fp:precise /MP /O2 /Ob2 /DNDEBUG

C flags (Debug): /DWIN32 /D_WINDOWS /W3 /D _CRT_SECURE_NO_DEPRECATE /D _CRT_NONSTDC_NO_DEPRECATE /D _SCL_SECURE_NO_WARNINGS /Gy /bigobj /Oi /fp:precise /MP /Zi /Ob0 /Od /RTC1

Linker flags (Release): /machine:x64 /INCREMENTAL:NO

Linker flags (Debug): /machine:x64 /debug /INCREMENTAL

ccache: NO

Precompiled headers: NO

Extra dependencies: cudart_static.lib nppc.lib nppial.lib nppicc.lib nppidei.lib nppif.lib nppig.lib nppim.lib nppist.lib nppisu.lib nppitc.lib npps.lib cublas.lib cudnn.lib cufft.lib -LIBPATH:C:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v12.5/lib/x64 -LIBPATH:C:/Program Files/NVIDIA GPU Computing Toolkit/CUDNN_9.2/lib/x64

3rdparty dependencies:

OpenCV modules:

To be built: aruco bgsegm bioinspired calib3d ccalib core cudaarithm cudabgsegm cudacodec cudafeatures2d cudafilters cudaimgproc cudalegacy cudaobjdetect cudaoptflow cudastereo cudawarping cudev datasets dnn dnn_objdetect dnn_superres dpm face features2d flann fuzzy gapi hfs highgui img_hash imgcodecs imgproc intensity_transform line_descriptor mcc ml objdetect optflow phase_unwrapping photo plot python3 quality rapid reg rgbd saliency shape signal stereo stitching structured_light superres surface_matching text tracking ts video videoio videostab wechat_qrcode world xfeatures2d ximgproc xobjdetect xphoto

Disabled: -

Disabled by dependency: -

Unavailable: alphamat cannops cvv freetype hdf java julia matlab ovis python2 python2 sfm viz

Applications: perf_tests apps

Documentation: NO

Non-free algorithms: YES

Windows RT support: NO

GUI:

QT: NO

Win32 UI: YES

OpenGL support: YES (opengl32 glu32)

VTK support: NO

Media I/O:

ZLib: build (ver 1.3.1)

JPEG: build-libjpeg-turbo (ver 3.0.3-70)

SIMD Support Request: YES

SIMD Support: NO

WEBP: build (ver encoder: 0x020f)

PNG: build (ver 1.6.43)

SIMD Support Request: YES

SIMD Support: YES (Intel SSE)

TIFF: build (ver 42 - 4.6.0)

JPEG 2000: build (ver 2.5.0)

OpenEXR: build (ver 2.3.0)

HDR: YES

SUNRASTER: YES

PXM: YES

PFM: YES

Video I/O:

DC1394: NO

FFMPEG: YES (prebuilt binaries)

avcodec: YES (58.134.100)

avformat: YES (58.76.100)

avutil: YES (56.70.100)

swscale: YES (5.9.100)

avresample: YES (4.0.0)

GStreamer: NO

DirectShow: YES

Media Foundation: YES

DXVA: YES

Parallel framework: Concurrency

Trace: YES (with Intel ITT)

Other third-party libraries:

Intel IPP: 2021.11.0 [2021.11.0]

at: D:/Program Files/opencv/opencv_cp312_410/opencv_410_cp312_amd64_bulid/3rdparty/ippicv/ippicv_win/icv

Intel IPP IW: sources (2021.11.0)

at: D:/Program Files/opencv/opencv_cp312_410/opencv_410_cp312_amd64_bulid/3rdparty/ippicv/ippicv_win/iw

Lapack: NO

Eigen: NO

Custom HAL: NO

Flatbuffers: builtin/3rdparty (23.5.9)

NVIDIA CUDA: YES (ver 12.5, CUFFT CUBLAS)

NVIDIA GPU arch: 89 90

NVIDIA PTX archs: 90

cuDNN: YES (ver 9.2.1)

OpenCL: YES (NVD3D11)

Include path: D:/Program Files/opencv/opencv_cp312_410/3rdparty/include/opencl/1.2

Link libraries: Dynamic load

Python 3:

Interpreter: C:/Program Files/Python312/python.exe (ver 3.12.4)

Libraries: optimized C:/Program Files/Python312/libs/python312.lib debug C:/Program Files/Python312/libs/python312_d.lib (ver 3.12.4)

Limited API: NO

numpy: C:/Program Files/Python312/Lib/site-packages/numpy/_core/include (ver 2.0.0)

install path: C:/Program Files/Python312/Lib/site-packages/cv2/python-3.12

Python (for build): C:/Program Files/Python312/python.exe

Java:

ant: NO

Java: NO

JNI: NO

Java wrappers: NO

Java tests: NO

Install to: D:/Program Files/opencv/opencv_cp312_410/opencv_410_cp312_amd64_bulid/install

-----------------------------------------------------------------

Configuring done (10.3s)

Generating done (13.0s)

四、使用VS Studio 进行编译

1、点击open project 自动打开vs studio

注意：该opencv_python3 是不可或缺的。没有的话检查是否勾选BUILD_opencv_python3

2、点击选择Release。

3、右键Install，点击生成

时间大概1个小时-2个小时左右，按cpu性能而定。若没有失败，则编译成功。

至此编译完成！！！

五、添加环境变量，修改config.py文件

1、修改config.py文件

编译完成后只有opencv/build/Install 文件夹（本文命名为opencv_cp312_410\opencv_410_cp312_amd64_bulid文件夹）是我们所需要的，将Install 文件夹移动到你所需的位置后，需要进行如下更改：

1、打开python 安装路径，或者在 cmake配置信息中取得，打开安装路径中的cv2文件夹（本文 install path 为C:/Program Files/Python312/Lib/site-packages/cv2/python-3.12，则打开C:/Program Files/Python312/Lib/site-packages/cv2 ）