cpp-OpenCV: C++ and OpenCV Setup (MinGW/GCC + CMake)

This repository contains C++ example code using the OpenCV library for computer vision tasks.

This guide details the steps to develop computer vision applications in C++ with OpenCV on Windows using the MinGW (GCC) compiler and CMake, managed from VS Code.

Documentation

OpenCV getting started guide

OpenCV official documentation

OpenCV C++ tutorials

OpenCV C++ API reference

🚀 Installation and Configuration

Prerequisites (Windows)

1. MinGW (GCC): Install a MinGW distribution (e.g., TDM-GCC or MinGW-w64) and ensure its bin directories are on the system Path.
2. CMake: Download and Install CMake at this link and choose the option to add it to the system Path.
3. VS Code: Install VS Code and the extensions: C/C++ and CMake Tools.

1. Build and Install OpenCV (MinGW-compatible)

To avoid linkage errors, build OpenCV from source using MinGW.

1. Download the Source: Download OpenCV source (e.g., 4.x) from the official site and extract to a location (example: C:\opt\opencv-4.12.0).

2. Configure and Build (terminal): Use CMake to generate MinGW Makefiles and then build and install:

   # Example configuration:
   cmake -S C:/opt/opencv-4.12.0 -B C:/opt/opencv_mingw_build -G "MinGW Makefiles" \
        -D CMAKE_BUILD_TYPE=Release \
        -D CMAKE_INSTALL_PREFIX="C:/opt/opencv_mingw" \
        -D BUILD_opencv_python3=OFF
   
   # Change to build directory
   cd C:/opt/opencv_mingw_build
   
   # Build and install (replace '12' with your CPU core count)
   mingw32-make -j12
   mingw32-make install

2. Configure PATH for OpenCV DLLs

Add the OpenCV DLL folder to your system PATH so executables can find runtime libraries:

1. Open Windows Environment Variables.

   Hit Win + R, type sysdm.cpl, go to the Advanced tab, and click Environment Variables.

2. Edit the user or system Path.

3. Add the DLL installation path (usually under bin), for example:

   C:\opt\opencv_mingw\x64\mingw\bin
   

   or in PowerShell:

   $env:PATH += ";C:\opt\opencv_mingw\x64\mingw\bin"

   to check the environment variable in the current session.

   echo $env:Path

Note: Close and reopen VS Code and terminals after updating PATH.

---

💻 Simple C++ Example with CMake

You can navigate to the FirstTrip folder for a simple example. Below are the steps to set up your own project.

1. CMakeLists.txt

Create this file at the project root. Set OpenCV_DIR to the folder containing OpenCVConfig.cmake.

cmake_minimum_required(VERSION 3.0)
project(cpp_OpenCV)

# Path to the OpenCV installation created by MinGW
# This should point to the folder that contains 'OpenCVConfig.cmake'.
set(OpenCV_DIR "C:/opt/opencv_mingw/lib/cmake/opencv4" CACHE PATH "Path to OpenCV MinGW install")

# Find OpenCV
find_package(OpenCV REQUIRED)

# Include OpenCV headers
include_directories(${OpenCV_INCLUDE_DIRS})

# Create executable
add_executable(cpp_OpenCV main.cpp)

# Link OpenCV libraries (MinGW-compatible .a/.dll)
target_link_libraries(cpp_OpenCV ${OpenCV_LIBS})

Create main.cpp with the following code (or use the version from FirstTrip): FirstTrip/main.cpp

Create main.cpp with the following code:

#include <opencv2/opencv.hpp>
#include <iostream>

const int WINDOW_HEIGHT = 600;
const int WINDOW_WIDTH = 800;

int main() {
    std::cout << "/* Starting */" << std::endl;
    cv::Mat image = cv::imread("blackhole.jpg");
    if (image.empty()) {
        std::cerr << "Error: Could not open image file." << std::endl;
        return -1;
    }

    // Create a window with fixed size (not resizable)
    cv::Mat resizedImage;
    cv::resize(image, resizedImage, cv::Size(WINDOW_WIDTH, WINDOW_HEIGHT));

    cv::namedWindow("Display Image", cv::WINDOW_NORMAL);
    cv::resizeWindow("Display Image", WINDOW_WIDTH, WINDOW_HEIGHT);

    cv::imshow("Display Image", resizedImage);
    cv::waitKey(0);
    return 0;
}

3. Configure the build with CMake

From your project directory:

mkdir build
cmake -B .\build\

Set the PATH for the current session (PowerShell):

$env:PATH += ";C:\opt\opencv_mingw\x64\mingw\bin"

To persist for the current user (run in an elevated PowerShell or reopen the session afterwards):

setx PATH "$($env:PATH);C:\opt\opencv_mingw\x64\mingw\bin"

After setting persistent PATH, close and reopen your terminal or IDE.

4. Build the project

Delete the existing build folder if it exists, then run the following commands to

Remove-Item -Recurse -Force build_debug, build_release;

Create debug and release build folders and build with MinGW Makefiles:

cmake -S . -B build_debug -G "MinGW Makefiles" -DCMAKE_BUILD_TYPE=Debug
cmake -S . -B build_release -G "MinGW Makefiles" -DCMAKE_BUILD_TYPE=Release

Debug build:

# Build Debug
cmake --build build_debug

# Run the executable
.\build_debug\cpp_OpenCV.exe

Release build:

# Build Release
cmake --build build_release

# Run the executable
.\build_release\cpp_OpenCV.exe