Как установить opencv на windows 10

Время на прочтение7 мин

Количество просмотров41K

Введение

В данной статье речь пойдет о сборке и установке OpenCV 4 для C/C++, Python 2 и Python 3 из исходных файлов с дополнительными модулями CUDA 10 на ОС Windows.

Я постарался включить сюда все тонкости и нюансы, с которыми можно столкнуться в ходе установки, и про которые не написано в официальном мануале.

Картинка с сайта: habr.com

Сборка тестировалась для:

• Windows 8.1 + Visual Studio 2017 + Python 2/3 + CUDA 10.0 + GeForce 840m
• Windows 10 + Visual Studio 2019 + Python 2/3 + CUDA 10.0 + GeForce GTX 1060

Внимание! Сборка не подойдет для версии OpenCV 4.0.1 и/или CUDA ниже 10-й версии. CUDA 9 и ниже поддерживается OpenCV 3.

Что потребуется для установки

В моей сборке использовались следующие инструменты:

1. CMake 3.15
2. MS Visual Studio 2019 64-бит + средства CMake С++ для Windows
3. Python 3.7.3 64-бит + NumPy 64-бит
4. Python 2.7.16 64-бит + NumPy 64-бит
5. CUDA 10.0
6. CuDNN 7.6.2
7. OpenCV 4.1.1 и OpenCV-contrib-4.1.1

Установка

Так как установка производится через консольные команды, то следует внимательно и аккуратно выполнять все шаги. Также, при необходимости, меняйте установочные пути на свои.
Для начала необходимо установить требуемое ПО, причем Visual Studio должна быть установлена до CUDA:

• CMake (версия >= 3.9.1)
• MS Visual Studio
• Python 3.7 (дистрибутив Anaconda3)
• CUDA 10.0
• CuDNN 7.6.2

После установки всех компонент убедитесь, что пути для CMake, Visual Studio, Python, CUDA, CuDNN прописаны в переменных PATH, PYTHONPATH, CUDA_PATH и cudnn соответственно.

Далее загружаем архивы исходников opencv-4.1.1 и opencv-contrib-4.1.1 в желаемое место (в моем случае это C:\OpenCV\).

git clone https://github.com/opencv/opencv.git -b "4.1.1"
git clone https://github.com/opencv/opencv_contrib.git -b "4.1.1"

Картинка с сайта: habr.com

Создаем папку build/ внутри opencv-4.1.1.

Картинка с сайта: habr.com

Далее сгенерируем файлы сборки с помощью cmake. Мы будем использовать консольный вариант cmake, так как cmake-gui путает типы некоторых переменных (например, OPENCV_PYTHON3_VERSION) и, как следствие, неправильно генерирует файлы.

Открываем консоль по пути C:\OpenCV\ и прописываем переменные.

set "opencvSource=opencv-4.1.1"
set "opencvExtraModules=opencv_contrib-4.1.1/modules"
set "opencvBuild=%opencvSource%\build"
set "compiler=Visual Studio 16 2019"
set "buildType=Release"

Примечание. Для Visual Studio 2017 генератор пишется как «Visual Studio 15 2017 Win64» и без флага -A.

Также можно явно указать питоновские библиотеки для python 2 и python 3 на случай, если сборщик не сможет найти их автоматически.

set "python2_executable=C:/Python27/python.exe"
set "python2_include_dir=C:/Python27/include"
set "python2_library=C:/Python27/libs/python27.lib"
set "python2_numpy_include_dirs=C:/Python27/Lib/site-packages/numpy/core/include"
set "python2_packages_path=C:/Python27/Lib/site-packages"

set "python3_executable=C:/Users/root/Anaconda3/python.exe"
set "python3_include_dir=C:/Users/root/Anaconda3/include"
set "python3_library=C:/Users/root/Anaconda3/libs/python37.lib"
set "python3_numpy_include_dirs=C:/Users/root/Anaconda3/lib/site-packages/numpy/core/include"
set "python3_packages_path=C:/Users/root/Anaconda3/Lib/site-packages"

Примечание. Обратите внимание на то, что библиотека NumPy должна быть той же разрядности, что и OpenCV. Проверить это легко:

import numpy.distutils.system_info as sysinfo
print(sysinfo.platform_bits)

Выполняем генерацию файлов сборки с помощью длинной команды ниже. При неудачной генерации или наличии ошибок после выполнения команды, повторную генерацию следует производить вычистив все файлы в build/ и .cache/.

cmake ^
-B"%opencvBuild%/" ^
-H"%opencvSource%/" ^
-G"%compiler%" ^
-Ax64 ^
-DCMAKE_BUILD_TYPE=%buildType% ^
-DBUILD_opencv_world=ON ^
-DINSTALL_TESTS=OFF ^
-DINSTALL_C_EXAMPLES=OFF ^
-DBUILD_EXAMPLES=OFF ^
-DOPENCV_EXTRA_MODULES_PATH="%opencvExtraModules%/" ^
-DBUILD_PROTOBUF=ON ^
-DBUILD_opencv_python_bindings_generator=ON ^
-DWITH_CUDA=ON ^
-DCUDA_FAST_MATH=ON ^
-DWITH_CUBLAS=ON ^
-DCUDA_ARCH_PTX=7.5 ^
-DBUILD_opencv_python2=ON ^
-DPYTHON2_EXECUTABLE="%python2_executable%" ^
-DOPENCV_PYTHON2_VERSION=2.7.16 ^
-DPYTHON2_INCLUDE_DIR="%python2_include_dir%" ^
-DPYTHON2_LIBRARY="%python2_library%" ^
-DPYTHON2_NUMPY_INCLUDE_DIRS="%python2_numpy_include_dirs%" ^
-DPYTHON2_PACKAGES_PATH="%python2_packages_path%" ^
-DBUILD_opencv_python3=ON ^
-DPYTHON3_EXECUTABLE="%python3_executable%" ^
-DOPENCV_PYTHON3_VERSION=3.7.3 ^
-DPYTHON3_INCLUDE_DIR="%python3_include_dir%" ^
-DPYTHON3_LIBRARY="%python3_library%" ^
-DPYTHON3_NUMPY_INCLUDE_DIRS="%python3_numpy_include_dirs%" ^
-DPYTHON3_PACKAGES_PATH="%python3_packages_path%"

Спустя примерно 10 минут в консоле должна появиться информация о сборке и завершающие строки «Configuring done» и «Generating done». Проверяем всю информацию, особенно разделы NVIDIA CUDA, Python 2, Python 3.

Картинка с сайта: habr.com

Далее собираем решение. На сборку может уйти несколько часов в зависимости от вашего процессора и версии Visual Studio.

cmake --build %opencvBuild% --target ALL_BUILD --config Release

Устанавливаем решение.

cmake --build %opencvBuild% --target INSTALL --config Release

После успешной установки создаем системную переменную OPENCV_DIR со значением C:\OpenCV\opencv-4.1.1\build\install\x64\vc15\bin и также добавляем ее в PATH.

Тест

Проверим работоспособность OpenCV с модулем CUDA на простом примере умножения матриц.

Пример на Python 3

import numpy as np
import cv2 as cv
import time

rand = np.random.random((1024, 1024)).astype(np.float32)
h_array1 = np.stack([rand, rand],axis=2)
h_array2 = h_array1
d_array1 = cv.cuda_GpuMat()
d_array2 = cv.cuda_GpuMat()
d_array1.upload(h_array1)
d_array2.upload(h_array2)

start = time.time()
cv.cuda.gemm(d_array1, d_array2, 1, None, 0, None, 1)
end = time.time()
print("Time elapsed:", end - start, "sec")

Вывод в консоли

Time elapsed: 0.3130002021789551

Пример на C++

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

using namespace std;
using namespace cv;
using namespace cv::cuda;

int main() {

	Mat h_array1 = Mat::ones(1024, 1024, CV_32FC2);
	Mat h_array2 = Mat::ones(1024, 1024, CV_32FC2);
	Mat h_array3 = Mat::zeros(1024, 1024, CV_32FC2);
	Mat h_result;
	GpuMat d_array1, d_array2, d_array3, d_result;

	d_array1.upload(h_array1);
	d_array2.upload(h_array2);

	const clock_t begin_time = clock();
	cuda::gemm(d_array1, d_array2, 1.0, d_array3, 0.0, d_result);
	cout << "Time elapsed: " << (float(clock() - begin_time) / CLOCKS_PER_SEC) << " sec" << endl;

	d_result.download(h_result);

	return 0;
}

Вывод в консоли

Time elapsed: 0.354

Удаление

Чтобы удалить OpenCV, нужно выполнить команду.

cmake --build %opencvBuild% --target uninstall --config Release

и удалить системную переменную OPENCV_DIR и убрать путь до OpenCV из PATH.

Заключение

В статье мы рассмотрели установку OpenCV 4 для ОС Windows 10. Данный алгоритм тестировался на Windows 8.1 и Windows 10, но, в теории, может собираться и на Windows 7. За дополнительной информацией можно обратиться к списку источников ниже.

P.S. Напоследок сведем все установочные команды в один .bat файл для тех, кто хочет автоматизировать установку.

cd C:\OpenCV

git clone https://github.com/opencv/opencv_contrib.git -b "4.1.1"
git clone https://github.com/opencv/opencv.git -b "4.1.1"

ren opencv opencv-4.1.1
ren opencv_contrib-4.1.1

set "opencvSource=opencv-4.1.1"
set "opencvExtraModules=opencv_contrib-4.1.1/modules"
set "opencvBuild=%opencvSource%\build"
set "compiler=Visual Studio 16 2019"
set "buildType=Release"

set "python2_executable=C:/Python27/python.exe"
set "python2_include_dir=C:/Python27/include"
set "python2_library=C:/Python27/libs/python27.lib"
set "python2_numpy_include_dirs=C:/Python27/Lib/site-packages/numpy/core/include"
set "python2_packages_path=C:/Python27/Lib/site-packages"

set "python3_executable=C:/Users/root/Anaconda3/python.exe"
set "python3_include_dir=C:/Users/root/Anaconda3/include"
set "python3_library=C:/Users/root/Anaconda3/libs/python37.lib"
set "python3_numpy_include_dirs=C:/Users/root/Anaconda3/lib/site-packages/numpy/core/include"
set "python3_packages_path=C:/Users/root/Anaconda3/Lib/site-packages"

cmake ^
-B"%opencvBuild%/" ^
-H"%opencvSource%/" ^
-G"%compiler%" ^
-Ax64 ^
-DCMAKE_BUILD_TYPE=%buildType% ^
-DBUILD_opencv_world=ON ^
-DINSTALL_TESTS=OFF ^
-DINSTALL_C_EXAMPLES=OFF ^
-DBUILD_EXAMPLES=OFF ^
-DOPENCV_EXTRA_MODULES_PATH="%opencvExtraModules%/" ^
-DBUILD_PROTOBUF=ON ^
-DBUILD_opencv_python_bindings_generator=ON ^
-DWITH_CUDA=ON ^
-DCUDA_FAST_MATH=ON ^
-DWITH_CUBLAS=ON ^
-DCUDA_ARCH_PTX=7.5 ^
-DBUILD_opencv_python2=ON ^
-DPYTHON2_EXECUTABLE="%python2_executable%" ^
-DOPENCV_PYTHON2_VERSION=2.7.16 ^
-DPYTHON2_INCLUDE_DIR="%python2_include_dir%" ^
-DPYTHON2_LIBRARY="%python2_library%" ^
-DPYTHON2_NUMPY_INCLUDE_DIRS="%python2_numpy_include_dirs%" ^
-DPYTHON2_PACKAGES_PATH="%python2_packages_path%" ^
-DBUILD_opencv_python3=ON ^
-DPYTHON3_EXECUTABLE="%python3_executable%" ^
-DOPENCV_PYTHON3_VERSION=3.7.3 ^
-DPYTHON3_INCLUDE_DIR="%python3_include_dir%" ^
-DPYTHON3_LIBRARY="%python3_library%" ^
-DPYTHON3_NUMPY_INCLUDE_DIRS="%python3_numpy_include_dirs%" ^
-DPYTHON3_PACKAGES_PATH="%python3_packages_path%"

cmake --build %opencvBuild% --target ALL_BUILD --config Release
cmake --build %opencvBuild% --target INSTALL --config Release

Список источников

1. docs.opencv.org/4.1.1/d3/d52/tutorial_windows_install.html
2. www.learnopencv.com/install-opencv-4-on-windows
3. jamesbowley.co.uk/build-opencv-4-0-0-with-cuda-10-0-and-intel-mkl-tbb-in-windows
4. lightbuzz.com/opencv-cuda
5. Глава «Building libraries from Source» книги Hands-on GPU-Accelerated Computer Vision with OpenCV and CUDA

• Clone this repository at <script src="https://gist.github.com/demid5111/6faa590e4fc5813550dd95cc1c538893.js"></script>

• Save demid5111/6faa590e4fc5813550dd95cc1c538893 to your computer and use it in GitHub Desktop.

Instructions for installing opencv on Windows 10

For developers wishing to explore computer vision, OpenCV is the best place to start. Installing OpenCV however is notoriously tricky. In this article we’ll step through a full build-from-source installation on Windows 10, including setting up Python and C++ development environments.

Building from Source

When installing OpenCV on Windows, we have two options. We can install pre-built binaries or perform our own build from source. The easiest way to get up and running quickly is to install the pre-built binaries. There are however a number of reasons why you may want to build from source:

1. You want to be able to debug OpenCV code and step into OpenCV functions. The pre-built Windows distribution contains binary libraries for release builds only therefore, if you want to debug, you need to build from source.
2. You want more control over the installation, for example which libraries and wrappers are included and which platforms are targeted.
3. You want to modify the OpenCV source code in your application.
4. You want to contribute to OpenCV development.

Installing the pre-built binaries is a great way to get started quickly however you will probably want to customize the install eventually as you get more familiar with what OpenCV has to offer. This article assumes you want to build from source and explains the full process on Windows 10.

Preparation

The first step is to download the source code. At the time of writing, the latest version of OpenCV is 4.2.0. The 4.2.0 source can be downloaded from OpenCV’s GitHub page. To have access to the latest OpenCV functionality, you will also want to download the contrib modules.

To build the OpenCV source code, a compiler is required. The suggested compiler for Windows is that provided by MSVC (Microsoft Visual C++). MSVC can either be installed as part of Microsoft Visual Studio or, if you don’t want the full Visual Studio install, you can just install the Visual Studio Build Tools. At the time of writing, the latest version of MSVC is version 16, which ships with Visual Studio 2019.

Unless you’re already familiar with C++, OpenCV’s Python interface is the best place to start. Regarding Python versions, the suggested version to use with OpenCV at the time of writing is Python 3.7. The more recent Python 3.8 has encountered issues with OpenCV bindings. Regarding target platforms, the default Python install available on www.python.org is x86 (32-bit). If you want to run x64 (64-bit) Python with OpenCV, the complication is that Numpy (a computational library which Python’s OpenCV bindings depend on) is only officially available in 32-bit format for Windows. The options therefore are as follows:

1. Go with 32-bit Python and the official 32-bit Numpy package. This is the easiest option for getting started. Once 32-bit Python is installed, running the following command at the command prompt installs Numpy at the system level: pip3 install numpy.
2. Use a 64-bit distribution of Python which includes 64-bit versions of supporting packages such as Numpy. Anaconda is the most common way of doing this. Distributions such as Anaconda are convenient as they have lots of tools readily available. The downside is you have less control over what is installed on your system and potentially many tools which you don’t need.
3. Build your own 64-bit version of Numpy. This is only suggested if you’re familiar with building Python packages from source and you absolutely want to run with 64-bit Python and don’t want to work with a 3rd party distribution such as Anaconda.

To facilitate building from source, OpenCV uses CMake. The CMake tool is a free and open source utility which simplifies the build process with a graphical interface. CMake can be obtained from the downloads section of the CMake website. Be sure to add CMake to your Path during the installation process to allow it to be called from the command prompt.

Building the Binaries

Now the preparation steps are complete, we can start the build process. Once the source files are downloaded and extracted, you should see two folders as per the image below.

Картинка с сайта: agmanic.com

Launch the CMake utility and set the path of the source location. A build folder needs to be designated which CMake will create automatically. Be careful to use forward slashes instead of backslashes, which reflects CMake’s Linux origins.

Картинка с сайта: agmanic.com

Click the ‘Configure’ button and accept the creation of the build directory. CMake will then prompt you to specify the generator. Attention needs to be paid here to x64 vs x86. Depending which architecture you’re targeting (discussed earlier in this article regarding Python), the appropriate selection needs to be made. Note that MSVC version 16 (Visual Studio 2019) defaults to x64 whilst MSVC 15 (Visual Studio 2017) defaults to x86. You may therefore need to override the default setting using the ‘Optional toolset to use’ field.

Картинка с сайта: agmanic.com

Upon clicking ‘Finish’, CMake will proceed to create the build configuration files in the designated build directory. Once the initial configuration step is complete, there are a number of build configuration changes to make in the red window of the CMake utility. The following changes are suggested:

1. Set CMAKE_INSTALL_PREFIX to the folder in which you want the binaries to be installed e.g. C:/Program Files/OpenCV
2. Check INSTALL_C_EXAMPLES and INSTALL_PYTHON_EXAMPLES
3. To have access to the full OpenCV functionality, check OPENCV_ENABLE_NONFREE
4. To include the contrib modules, set OPENCV_EXTRA_MODULES_PATH to the contrib modules source location e.g. C:\Users\[user]\Downloads\opencv_contrib-4.2.0\opencv_contrib-4.2.0\modules
5. Set PYTHON3_PACKAGES_PATH to where you want the Python bindings installed – by default this will be C:/Users/[user]/AppData/Local/Programs/Python/Python37-32/Lib/site-packages

Once the above changes have been completed, click ‘Configure’ again to update the build configuration files then click ‘Generate’ to generate the build files.

Once the build files are generated, we can proceed with the build itself. Open a command prompt, navigate to the build directory which contains CMake’s output and run the following command: cmake.exe –build . –config Release –target INSTALL. The command prompt may need to be opened with administrator privileges, depending on where the binaries are being installed. This process will take approximately 15 minutes, depending on your system. When complete, the binaries will be installed in the folder indicated by the CMAKE_INSTALL_PREFIX field during CMake configuration. If you want to add debug functionality, also run the following command in the build directory: cmake.exe –build . –config Debug –target INSTALL. To complete the build process, add the location of the binaries to your Path e.g. C:\Program Files\OpenCV\x64\vc16\bin (this will depend on the location indicated by the CMAKE_INSTALL_PREFIX field during CMake configuration.

Testing the Python bindings

If the PYTHON3_PACKAGES_PATH field in CMake was set to the system site packages folder (see CMake configuration section earlier in this article), OpenCV’s Python bindings should work without any further configuration. Launch the Python interpreter, paying attention to x64 vs x86 if you have multiple versions installed, and enter ‘import cv2’. If successful, you will receive no error messages and be able to check the version.

Картинка с сайта: agmanic.com

If you have errors when trying to import cv2, the most likely causes are either x64 vs x86 compatibility or the cv2 bindings not being on Python’s path. The x64 vs x86 compatibility issues are discussed earlier in this article: Python, Numpy and the compiler need to be aligned on the same architecture. Regarding Python’s path, the cv2 module may need to be added to the PYTHONPATH environment variable if the cv2 module was installed in a location other that the system site packages folder.

Setting up a C++ development environment

On Windows, the most common C++ development environment is Microsoft Visual Studio therefore we’ll focus on Visual Studio for this article. Once the compiled binaries are correctly installed, a development environment such as Visual Studio requires the paths to the include files and library files in order to work with an external library such as OpenCV.

The first step is to create an environment variable pointing to the install directory of the compiled binaries. This simplifies the addition of paths to Visual Studio and avoids having to modify the Visual Studio project from one machine to another. For example, on my machine, I created an environment variable called OPENCVDIR, pointing to C:\Program Files\OpenCV.

Within your Visual Studio C++ project, open the Project Properties dialogue and navigate to C/C++ > General. Under ‘Additional Include Directories’, add the environment variable created earlier, followed by the path to the include files.

Картинка с сайта: agmanic.com

The next step is to add the path to the libraries for linking. Again within the Project Properties dialogue, navigate to Linker > General and add the path to the library files under ‘Additional Library Directories’.

Картинка с сайта: agmanic.com

The final step is to add the specific library files to be linked under the Linker > Input section. To start with, it is suggested to add opencv_core420.lib, opencv_highgui420.lib, opencv_imgcodecs420.lib and opencv_imgproc420.lib. You can add more libraries later as you explore more of the OpenCV functionality.

Картинка с сайта: agmanic.com

Depending on whether you want to link statically or dynamically, a final step is to set the Runtime Library property within Project Properties > Configuration Properties > C/C++ > Code Generation. Your OpenCV C++ development environment is now setup and ready for building OpenCV applications.