周六日在家捣腾了一下，把过程记录下来。

前置条件

1. Visual Studio C++ 生成工具
2. 和本机显卡适配的CUDA
3. 与CUDA匹配的cuDNN
4. Python 3
5. NumPy
6. OpenCV源代码以及对应版本的OpenCV-contrib模块源码
7. CMake

Visual Studio

下载Visual Studio（我本机的是VS2022），通过Visual Studio Installer安装程序，安装C++工具集（或C++工作负荷），详细安装过程可参考这里。

CUDA和cuDNN

下载安装最新版的CUDA Toolkit，注意与本地GPU兼容，或者检查本地路径，看是否已经安装CUDA工具包。以我本机为例，CUDA12.5安装在C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v12.5。同上，登录NVIDIA账号下载cuDNN，并将cuDNN文件中的内容复制到CUDA Toolkit所在目录（如C:\Program Files\NVIDIA\CUDNN\vX.X）的bin、include和lib/x64等文件夹下，我本机的cuDNN是9.2.1版。

Python、NumPy及pip

安装Python3.x版本，由于需要使用numpy矩阵替代cv:Mat,，故还需安装numpy，保证已经安装好numpy（pip install numpy）并确保包括opencv-python和opencv-contrib-python等opencv包卸载干净。

pip uninstall opencv-python

pip uninstall opencv-contrib-python

删除cv2目录——YOUR_PYTHON_PATH/Lib/site-packages/cv2

OpenCV

从github仓库下载，或克隆仓库到本地，内容包括OpenCV及版本匹配的opencv-contrib。

CMake配置

给opencv和opencv-contrib创建build目录，然后配置cmake。Cmake配置可参考官网链接：OpenCV: OpenCV configuration options reference。

这是一个漫长的过程，中途需要下载3rdparty文件夹里引用的第三方内容，个别库还可能出错，需要手工下载。

本例我们把Python也选上：

General configuration for OpenCV 4.10.0 =====================================Version control: unknownPlatform:Timestamp: 2024-07-20T06:31:04ZHost: Windows 10.0.22631 AMD64CMake: 3.29.0CMake generator: Visual Studio 17 2022CMake build tool: C:/Program Files/Microsoft Visual Studio/2022/Enterprise/MSBuild/Current/Bin/amd64/MSBuild.exeMSVC: 1940Configuration: Debug ReleaseCPU/HW features:Baseline: SSE SSE2 SSE3requested: SSE3Dispatched code generation: SSE4_1 SSE4_2 FP16 AVX AVX2 AVX512_SKXrequested: SSE4_1 SSE4_2 AVX FP16 AVX2 AVX512_SKXSSE4_1 (18 files): + SSSE3 SSE4_1SSE4_2 (2 files): + SSSE3 SSE4_1 POPCNT SSE4_2FP16 (1 files): + SSSE3 SSE4_1 POPCNT SSE4_2 FP16 AVXAVX (9 files): + SSSE3 SSE4_1 POPCNT SSE4_2 AVXAVX2 (38 files): + SSSE3 SSE4_1 POPCNT SSE4_2 FP16 FMA3 AVX AVX2AVX512_SKX (8 files): + SSSE3 SSE4_1 POPCNT SSE4_2 FP16 FMA3 AVX AVX2 AVX_512F AVX512_COMMON AVX512_SKXC/C++:Built as dynamic libs?: YESC++ standard: 11C++ Compiler: C:/Program Files/Microsoft Visual Studio/2022/Enterprise/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 /DNDEBUGC++ 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 /RTC1C Compiler: C:/Program Files/Microsoft Visual Studio/2022/Enterprise/VC/Tools/MSVC/14.40.33807/bin/Hostx64/x64/cl.exeC 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 /DNDEBUGC 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 /RTC1Linker flags (Release): /machine:x64 /INCREMENTAL:NOLinker flags (Debug): /machine:x64 /debug /INCREMENTALccache: NOPrecompiled headers: YESExtra dependencies:3rdparty dependencies:OpenCV modules:To be built: calib3d core dnn features2d flann gapi highgui imgcodecs imgproc java ml objdetect photo python3 stitching ts video videoioDisabled: worldDisabled by dependency: -Unavailable: python2Applications: tests perf_tests appsDocumentation: NONon-free algorithms: NOWindows RT support: NOGUI: WIN32UIWin32 UI: YESVTK support: NOMedia I/O:ZLib: build (ver 1.3.1)JPEG: build-libjpeg-turbo (ver 3.0.3-70)SIMD Support Request: YESSIMD Support: NOWEBP: build (ver encoder: 0x020f)PNG: build (ver 1.6.43)SIMD Support Request: YESSIMD 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: YESSUNRASTER: YESPXM: YESPFM: YESVideo I/O:DC1394: NOFFMPEG: 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: NODirectShow: YESMedia Foundation: YESDXVA: YESParallel framework: ConcurrencyTrace: YES (with Intel ITT)Other third-party libraries:Intel IPP: 2021.11.0 [2021.11.0]at: D:/Data/source/collection/OpenCV/4100/build/3rdparty/ippicv/ippicv_win/icvIntel IPP IW: sources (2021.11.0)at: D:/Data/source/collection/OpenCV/4100/build/3rdparty/ippicv/ippicv_win/iwLapack: NOEigen: NOCustom HAL: NOProtobuf: build (3.19.1)Flatbuffers: builtin/3rdparty (23.5.9)OpenCL: YES (NVD3D11)Include path: D:/Data/source/collection/OpenCV/4100/opencv-4.10.0/3rdparty/include/opencl/1.2Link libraries: Dynamic loadPython 3:Interpreter: D:/miniconda3/python.exe (ver 3.11.7)Libraries: D:/miniconda3/libs/python311.lib (ver 3.11.7)Limited API: NOnumpy: D:/miniconda3/Lib/site-packages/numpy/core/include (ver 1.26.1)install path: D:/miniconda3/Lib/site-packages/cv2/python-3.11Python (for build): D:/miniconda3/python.exeJava:ant: NOJava: YES (ver 1.8.0.371)JNI: D:/Program Files/jdk-1.8/include D:/Program Files/jdk-1.8/include/win32 D:/Program Files/jdk-1.8/includeJava wrappers: YES (JAVA)Java tests: NOInstall to: D:/Data/source/collection/OpenCV/4100/build/install-----------------------------------------------------------------Configuring done (136.2s)

配置后好再修改以下两个参数，其中CUDA_ARCH_BIN找到CUDA Toolkit后，目前的版本会自动选上。

之前还生成了VTK，故加上了VTK路径（这是VTK的cmake生成路径）：

点击“生成”按钮，生成VS工程。

从图可见，Java和Python的绑定工程都有了。

Visual Studio生成

执行ALL_BUILD生成工程，执行INSTALL进行安装。

生成和安装是一个漫长的等待……

注：要直接安装到Python环境中，需要用管理员身份打开VS，然后生成INSTALL项目。

安装效果及应用

#pragma comment(lib, "opencv_core4100.lib")#include <iostream>
#include <opencv2/core/cuda.hpp>int main()
{int deviceCount = cv::cuda::getCudaEnabledDeviceCount();std::cout << "CUDA Device Number: " << deviceCount << std::endl;cv::cuda::printCudaDeviceInfo(0);
}

Python 3.11.7 | packaged by Anaconda, Inc. | (main, Dec 15 2023, 18:05:47) [MSC v.1916 64 bit (AMD64)] on win32Type "help", "copyright", "credits" or "license" for more information.>>> import cv2>>> print(cv2.__version__)4.10.0>>> print(cv2.cuda.getCudaEnabledDeviceCount())1>>> cv2.cuda.printCudaDeviceInfo(0)*** CUDA Device Query (Runtime API) version (CUDART static linking) ***Device count: 1Device 0: "NVIDIA GeForce RTX 3070 Ti Laptop GPU"CUDA Driver Version / Runtime Version          12.50 / 12.50CUDA Capability Major/Minor version number:    8.6Total amount of global memory:                 8192 MBytes (8589410304 bytes)GPU Clock Speed:                               1.41 GHzMax Texture Dimension Size (x,y,z)             1D=(131072), 2D=(131072,65536), 3D=(16384,16384,16384)Max Layered Texture Size (dim) x layers        1D=(32768) x 2048, 2D=(32768,32768) x 2048Total amount of constant memory:               65536 bytesTotal amount of shared memory per block:       49152 bytesTotal number of registers available per block: 65536Warp size:                                     32Maximum number of threads per block:           1024Maximum sizes of each dimension of a block:    1024 x 1024 x 64Maximum sizes of each dimension of a grid:     2147483647 x 65535 x 65535Maximum memory pitch:                          2147483647 bytesTexture alignment:                             512 bytesConcurrent copy and execution:                 Yes with 1 copy engine(s)Run time limit on kernels:                     YesIntegrated GPU sharing Host Memory:            NoSupport host page-locked memory mapping:       YesConcurrent kernel execution:                   YesAlignment requirement for Surfaces:            YesDevice has ECC support enabled:                NoDevice is using TCC driver mode:               NoDevice supports Unified Addressing (UVA):      YesDevice PCI Bus ID / PCI location ID:           1 / 0Compute Mode:Default (multiple host threads can use ::cudaSetDevice() with device simultaneously)deviceQuery, CUDA Driver = CUDART, CUDA Driver Version  = 12.50, CUDA Runtime Version = 12.50, NumDevs = 1

遇到的几个问题

1. Visual Studio 已安装的Python版本影响

之前安装VS2022时装了Python开发负荷（Python3.9），导致cmake的时候绑死了该环境，且指向conda里的Python环境，其Libraries还是指向3.9。卸载VS中的Python可以解决。

    2. 缺Nvidia Video Codec SDK导致的警告

CMake Warning at D:/Data/source/collection/OpenCV/4100/opencv_contrib-4.10.0/modules/cudacodec/CMakeLists.txt:26 (message):
cudacodec::VideoReader requires Nvidia Video Codec SDK. Please resolve
dependency or disable WITH_NVCUVID=OFF

CMake Warning at D:/Data/source/collection/OpenCV/4100/opencv_contrib-4.10.0/modules/cudacodec/CMakeLists.txt:30 (message):
cudacodec::VideoWriter requires Nvidia Video Codec SDK. Please resolve
dependency or disable WITH_NVCUVENC=OFF

下载nvidia Video Codec SDK，并把lib和头文件（interface目录）分别复制到cuda toolkit的lib/x64和include目录，问题解决。

    3. CUDA版本问题导致的错误CMake Error at cmake/OpenCVDetectCUDAUtils.cmake :297 (list)  list GET given empty list

因我的Visual Studio是17.10.4，在CUDA12.2上构建，则会出现这个问题，因为根据官方文档，CUDA Toolkit 12.2 只支持到17.0的Visual Studio，如下图：

CUDA Installation Guide Microsoft Windows (nvidia.com)

更换为CUDA 12.5，可以解决这个问题：

https://docs.nvidia.com/cuda/archive/12.5.0/cuda-installation-guide-microsoft-windows/index.html#system-requirements

    5. Python使用CV2时dll缺失错误

ImportError: DLL load failed while importing cv2: 找不到指定的模块。

Traceback (most recent call last):

  File "<stdin>", line 1, in <module>

  File "D:\miniconda3\Lib\site-packages\cv2\__init__.py", line 181, in <module>

    bootstrap()

  File "D:\miniconda3\Lib\site-packages\cv2\__init__.py", line 153, in bootstrap

    native_module = importlib.import_module("cv2")

                    ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

  File "D:\miniconda3\Lib\importlib\__init__.py", line 126, in import_module

    return _bootstrap._gcd_import(name[level:], package, level)

提示dll缺失，使用ProcessMonitor，添加python.exe过滤器，重现错误，追出出错原因：

发现原来是自己编译VTK带来的锅，自己搞的VTK，含着泪也要把它搞定，So，加到cv2的config.py中，但导致别的错误（都怪自己，把VTK的debug版和release版放一起了），单独抽取当中的release版，加入到环境变量或cv2的config.py，或者直接拷贝到site-packages->cv2->python-3.11目录。搞定，问题解决。

参考资料

Quick and Easy OpenCV Python Installation with Cuda GPU in Under 10 Minutes (youtube.com)

GitHub - chrismeunier/OpenCV-CUDA-installation: Saving the process to install OpenCV for Python 3 with CUDA bindings

Unable to enable Cudacodec VideoReader · Issue #11220 · opencv/opencv · GitHub

OpenCV: OpenCV configuration options reference

CUDA Installation Guide for Microsoft Windows (nvidia.com)

windows10+VS2022编译安装opencv-python_重新编译opencv-python-CSDN博客