This guide covers configuring OpenCV from the official pre-built Windows binary. For a CUDA build see CUDA / GPU guide.

Step 1 — Download

Download the official OpenCV Windows binary

Download opencv-4.12.0-windows.exe from the download page or from github.com/opencv/opencv/releases. Run the self-extracting archive and extract to C:\opencv.

After extraction, the key folders are:

C:\opencv\build\include\ # header files

C:\opencv\build\x64\vc16\lib\ # .lib files for linking

C:\opencv\build\x64\vc16\bin\ # .dll files — add to PATH

Step 2 — PATH

Add OpenCV bin to Windows PATH

Before configuring Visual Studio, add the OpenCV bin directory to your system PATH so Windows can find the DLLs at runtime:

1
Open Environment Variables
Win+S → search Edit the system environment variables → click Environment Variables.
2
Edit System Path
Under System variables, click Path → Edit → New.
3
Add the bin path

Path to add
C:\opencv\build\x64\vc16\bin

Click OK on all dialogs. Restart Visual Studio for the change to take effect.

Step 3 — Visual Studio

Configure Visual Studio project

For each Visual Studio project that uses OpenCV, set these three properties. Right-click your project → Properties.

Property page	Setting	Value
C/C++ → General	Additional Include Directories	`C:\opencv\build\include`
Linker → General	Additional Library Directories	`C:\opencv\build\x64\vc16\lib`
Linker → Input	Additional Dependencies	`opencv_world4120.lib` (Release) or `opencv_world4120d.lib` (Debug)

Make sure your project platform is set to x64, not Win32. OpenCV pre-built binaries are 64-bit only.

Test it

Verify the setup with a minimal test

main.cpp

#include <opencv2/opencv.hpp>

#include <iostream>

int main() {

std::cout << "OpenCV: " << CV_VERSION << std::endl;

return 0;

}

// Expected output:

OpenCV: 4.12.0

FAQ

C++ setup questions

opencv_world4120.dll was not found

The DLL directory is not on PATH. Add C:\opencv\build\x64\vc16\bin to your system PATH and restart Visual Studio and your terminal. See opencv_world.dll fix guide.

LNK1107: invalid or corrupt file: cannot read

You are linking against the wrong .lib file. For Release builds use opencv_world4120.lib. For Debug builds use opencv_world4120d.lib. Make sure your project configuration (Release/Debug) matches the lib file.

Which vc version to use (vc14, vc15, vc16)?

Use the version matching your Visual Studio: vc14 = VS 2015, vc15 = VS 2017, vc16 = VS 2019/2022. OpenCV 4.12.0 ships with vc16 which works for both VS 2019 and VS 2022.

Can I use CMake instead of Visual Studio project properties?

Yes — see the CMake guide. CMake is recommended for projects that need to be portable or cross-platform, or when building multiple executables from one configuration.

Related guides

opencv_world.dll fix →