mkdir build-arm64 && cd build-arm64
cmake -DPRINT_DEBUG=ON -DBUILD_TESTING=ON -DCMAKE_OSX_ARCHITECTURES=arm64 \
-DDEHANCER_TARGET_ARCH=arm64-apple-macos11 -DDEHANCER_GPU_OPENCL=ON \
-DDEHANCER_GPU_METAL=OFF -DDEHANCER_GPU_CUDA=OFF ..
mkdir build-x86_64 && cd build-x86_64
cmake -DPRINT_DEBUG=ON -DBUILD_TESTING=ON \
-DCMAKE_OSX_ARCHITECTURES=x86_64 -DDEHANCER_TARGET_ARCH=x86_64-apple-macos10.14 \
-DDEHANCER_GPU_OPENCL=OFF -DDEHANCER_GPU_METAL=ON -DDEHANCER_GPU_CUDA=OFF ..
mkdir build-arm64-ios && cd build-arm64-ios
export PKG_CONFIG_PATH=~/Develop/local/ios-debug/dehancer/lib/pkgconfig
cmake -G Xcode \
-DCMAKE_TOOLCHAIN_FILE=~/Develop/Dehancer/Dehancer-Plugins/ios-cmake/ios.toolchain.cmake \
-DPLATFORM=OS64COMBINED \
-DDEPLOYMENT_TARGET=13.0 \
-DENABLE_BITCODE=ON \
-DBUILD_TESTING=OFF \
-DCMAKE_INSTALL_PREFIX=~/Develop/local/ios-debug/dehancer \
-DOPENCV_FRAMEWORK_PATH=~/Develop/local/ios-debug/lib \
-DOPENCV_INCLUDES_PATH=~/Develop/local/ios-debug/include \
-DENABLE_ARC=OFF \
-DDEHANCER_GPU_METAL=ON \
-DDEHANCER_GPU_OPENCL=OFF \
-DDEHANCER_GPU_CUDA=OFF \
-DUSE_OPENCOLORIO=OFF \
..
cmake --build . --config Debug --target EmbeddedWatermarks_metal
cmake --build . --config Debug && cmake --install . --config Debug
lipo -create libdehancer_gpulib_metal_iphoneos.a libdehancer_gpulib_metal_iphonesimulator.a -output libdehancer_gpulib_metal.a
# CUDA
cmake -G
"Ninja"
-DCMAKE_VERBOSE_MAKEFILE=ON
-DDEHANCER_GPU_CUDA=ON
-DDEHANCER_GPU_OPENCL=OFF
-DBUILD_TESTING=ON
-DPRINT_DEBUG=ON
-DCMAKE_TOOLCHAIN_FILE=C:/vcpkg/scripts/buildsystems/vcpkg.cmake
-DVCPKG_TARGET_TRIPLET=x64-windows-static
..
Nasm:
brew install nasm
or
yum install nasm
1. xxd from msys2
2. cuda toolkit:
https://developer.nvidia.com/cuda-10.2-download-archive?target_os=Windows&target_arch=x86_64&target_version=10&target_type=exelocal
git clone https://github.com/libjpeg-turbo/libjpeg-turbo
cd libjpeg-turbo && mkdir build && cd build
cmake -DENABLE_SHARED=OFF -DENABLE_STATIC=ON -DCMAKE_POSITION_INDEPENDENT_CODE=OFF -DCMAKE_C_FLAGS="-fPIC" -DCMAKE_BUILD_TYPE=Release -DCMAKE_INSTALL_PREFIX=/usr/local ..
make -j4 && make install
yum install llvm-toolset-7
scl enable llvm-toolset-7 bash
cmake -DCMAKE_C_COMPILER=clang -DCMAKE_CXX_COMPILER=clang++ -DBUILD_TESTING=ON ..
# add deb http://apt.llvm.org/hirsute/ llvm-toolchain-hirsute-11 main
sudo apt-key adv --keyserver keyserver.ubuntu.com --recv-keys 15CF4D18AF4F7421
sudo apt update
sudo apt upgrade
sudo apt install clang-11 clang-tools-11 clang-11-doc libclang-common-11-dev libclang-11-dev libclang1-11 clang-format-11 clangd-11
sudo apt-get install libllvm-11-ocaml-dev libllvm11 llvm-11 llvm-11-dev llvm-11-doc llvm-11-examples llvm-11-runtime
sudo apt-get install lldb-11
sudo apt-get install lld-11
sudo apt-get install libc++-11-dev libc++abi-11-dev
sudo apt-get install libomp-11-dev
sudo ln -s /usr/bin/clang-11 /usr/bin/clang
sudo ln -s /usr/bin/clang++-11 /usr/bin/clang++
sudo apt install libssl1.1 ocl-icd-opencl-dev fakeroot xorriso
sudo apt install gfortran
sudo aptitude install liblapack-dev
git clone https://github.com/Reference-LAPACK/lapack-release.git
cd lapack-release && mkdir build && cd build
cmake -DBUILD_SHARED_LIBS=OFF -DCMAKE_C_COMPILER=clang ..
make -j6 && sudo make install
#
# MacOS Universal binary
#
scripts/build_opencv_macos.sh [--prefix <installation path>]
... progress ...
Password: <type admin password>
use -DOpenCV_DIR=/usr/local/universal by default or -DOpenCV_DIR=<installation path>
or
git clone -b 4.x https://github.com/opencv/opencv.git
cd opencv
cmake -DCMAKE_OSX_ARCHITECTURES="arm64;x86_64" -DCMAKE_OSX_DEPLOYMENT_TARGET=11.0 -DCMAKE_INSTALL_PREFIX:PATH=/usr/local/universal \
-DBUILD_SHARED_LIBS=OFF -DWITH_FFMPEG=OFF -DWITH_V4L=OFF -DWITH_EIGEN=OFF\
-DWITH_JPEG=ON -DBUILD_JPEG=ON \
-DVIDEOIO_ENABLE_PLUGINS=ON -DOPENCV_GENERATE_PKGCONFIG=ON \
-DBUILD_EXAMPLES=OFF -DBUILD_TESTS=OFF -DCMAKE_FIND_LIBRARY_SUFFIXES=".a"\
-DOPENCV_DNN_OPENCL=OFF -DCAROTENE_NEON_ARCH=OFF -DBUILD_opencv_dnn=OFF \
-DWITH_PROTOBUF=OFF -DBUILD_PROTOBUF=OFF ..
cmake --build . -j12 --config Release
sudo cmake --build . -j12 --target install
# To make opencv on M1 for x86 copy Terminal.app to Intel Terminal.app
# open "Get Info", choose "Open using Rosetta"
#
################################################################
#
# On mode i386:
# $env /usr/bin/arch -x86_64 /bin/zsh
# Check mode
# $ arch
#
################################################################
# mkdir build_opencv_x86_64 && cd build_opencv_x86_64
# cmake -DCMAKE_OSX_ARCHITECTURES=x86_64 -DCMAKE_INSTALL_PREFIX:PATH=/usr/local/x86_64
# ...
mkdir build_opencv_arm64 && cd build_opencv_arm64
# M2 Macos Ventura
arch -arm64 cmake -DCMAKE_OSX_ARCHITECTURES=arm64 -DCMAKE_OSX_DEPLOYMENT_TARGET=11.0 -DCMAKE_OSX_SYSROOT=macosx -DCMAKE_INSTALL_PREFIX:PATH=/usr/local/arm64 \
-DBUILD_SHARED_LIBS=OFF -DWITH_FFMPEG=OFF -DWITH_V4L=OFF -DWITH_EIGEN=OFF\
-DWITH_JPEG=ON -DBUILD_JPEG=ON -DCMAKE_OSX_SYSROOT=macosx \
-DVIDEOIO_ENABLE_PLUGINS=ON -DOPENCV_GENERATE_PKGCONFIG=ON \
-DBUILD_EXAMPLES=OFF -DBUILD_TESTS=OFF -DCMAKE_FIND_LIBRARY_SUFFIXES=".a" ..
arch -arm64 cmake --build . -j16 --config Release
sudo cmake --install . --config Release
M1
cmake -DCMAKE_OSX_ARCHITECTURES=arm64 -DCMAKE_INSTALL_PREFIX:PATH=/usr/local/arm64 \
-DBUILD_SHARED_LIBS=OFF -DWITH_FFMPEG=OFF -DWITH_V4L=OFF -DWITH_EIGEN=ON\
-DWITH_JPEG=ON -DBUILD_JPEG=OFF \
-DJPEG_INCLUDE_DIR=/usr/local/jpeg-turbo/include \
-DJPEG_LIBRARY=/usr/local/jpeg-turbo/lib/libjpeg.a \
-DVIDEOIO_ENABLE_PLUGINS=OFF -DOPENCV_GENERATE_PKGCONFIG=ON \
-DBUILD_EXAMPLES=OFF -DBUILD_TESTS=OFF -DCMAKE_FIND_LIBRARY_SUFFIXES=".a" ..
$env /usr/bin/arch -x86_64 /bin/zsh
mkdir build_opencv_x86_64 && cd build_opencv_x86_64
cmake -DCMAKE_OSX_ARCHITECTURES=x86_64 -DCMAKE_OSX_DEPLOYMENT_TARGET=10.14 \
-DCMAKE_INSTALL_PREFIX:PATH=/usr/local/x86_64 \
-DWITH_JPEG=ON -DBUILD_JPEG=OFF \
-DJPEG_INCLUDE_DIR=/usr/local/jpeg-turbo/include \
-DJPEG_LIBRARY=/usr/local/jpeg-turbo/lib/libjpeg.a \
-DBUILD_SHARED_LIBS=OFF -DWITH_FFMPEG=OFF -DWITH_V4L=OFF -DVIDEOIO_ENABLE_PLUGINS=OFF -DWITH_EIGEN=ON\
-DOPENCV_GENERATE_PKGCONFIG=ON DBUILD_EXAMPLES=OFF -DBUILD_TESTS=OFF \
-DCMAKE_FIND_LIBRARY_SUFFIXES=".a" ..
# on centos add -DCMAKE_C_COMPILER=clang -DCMAKE_CXX_COMPILER=clang++
make -j7 && make install
# Centos 8
sudo dnf install epel-release dnf-utils
sudo yum-config-manager --add-repo=https://negativo17.org/repos/epel-multimedia.repo
sudo dnf install ffmpeg
sudo dnf install ffmpeg-devel
# Ubuntu 20.04
sudo apt-get install libavcodec-dev libavformat-dev libavutil-dev libswscale-dev libavresample-dev
mkdir build_opencv_x86_64 && cd build_opencv_x86_64
cmake -DWITH_CUDA=OFF -DWITH_OPENCL=ON -DWITH_OPENGL=ON -DWITH_V4L=ON -DBUILD_SHARED_LIBS=OFF -DWITH_EIGEN=OFF\
-DWITH_FFMPEG=ON -DVIDEOIO_ENABLE_PLUGINS=ON -DOPENCV_GENERATE_PKGCONFIG=ON -DBUILD_EXAMPLES=OFF\
-DBUILD_TESTS=OFF -DBUILD_opencv_java=OFF -DCMAKE_FIND_LIBRARY_SUFFIXES=".a" ..
# -DVIDEOIO_ENABLE_PLUGINS=ON can be OFF
# only plugin and Centos:
cmake -DWITH_CUDA=OFF -DWITH_OPENCL=ON -DWITH_OPENGL=ON -DWITH_V4L=ON -DBUILD_SHARED_LIBS=OFF -DWITH_EIGEN=OFF \
-DWITH_FFMPEG=ON -DVIDEOIO_ENABLE_PLUGINS=OFF -DOPENCV_GENERATE_PKGCONFIG=ON -DBUILD_EXAMPLES=OFF\
-DBUILD_TESTS=OFF -DBUILD_opencv_java=OFF -DCMAKE_FIND_LIBRARY_SUFFIXES=".a"\
-DCMAKE_C_COMPILER=clang -DCMAKE_CXX_COMPILER=clang++ -DCMAKE_INSTALL_PREFIX=/usr/local\
-DBUILD_LIST=core,imgcodecs,imgproc,photo,video,videoio ..
# Windows 10 (if you want build from sources)
export PATH=$PATH:/c/vcpkg/downloads/tools/cmake-3.21.1-windows/cmake-3.21.1-windows-i386/bin
cmake -Ax64 -DWITH_CUDA=OFF -DWITH_OPENCL=ON -DWITH_OPENGL=ON -DWITH_V4L=ON -DBUILD_SHARED_LIBS=OFF \
-DWITH_EIGEN=OFF -DWITH_FFMPEG=ON -DVIDEOIO_ENABLE_PLUGINS=ON -DOPENCV_GENERATE_PKGCONFIG=ON \
-DBUILD_EXAMPLES=OFF -DBUILD_TESTS=OFF -DBUILD_opencv_java=OFF -DCMAKE_FIND_LIBRARY_SUFFIXES=".a" \
-DCMAKE_TOOLCHAIN_FILE=C:/vcpkg/scripts/buildsystems/vcpkg.cmake -DVCPKG_TARGET_TRIPLET=x64-windows-static \
"-DCMAKE_C_COMPILER=C:/Program Files (x86)/Microsoft Visual Studio/2019/Community/VC/Tools/Llvm/x64/bin/clang-cl.exe" \
"-DCMAKE_CXX_COMPILER=C:/Program Files (x86)/Microsoft Visual Studio/2019/Community/VC/Tools/Llvm/x64/bin/clang-cl.exe" \
-DCMAKE_INSTALL_PREFIX=C:/Users/denn/Dehancer/local/dehancer -DCMAKE_BUILD_TYPE=Release ..
cmake --build . -j12 --target ALL_BUILD --config Release
cmake --install . --config Release
wget https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2004/x86_64/cuda-ubuntu2004.pin
sudo mv cuda-ubuntu2004.pin /etc/apt/preferences.d/cuda-repository-pin-600
sudo apt-key adv --fetch-keys https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2004/x86_64/7fa2af80.pub
sudo add-apt-repository "deb https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2004/x86_64/ /"
sudo apt-get update
sudo apt-get -y install cuda
- PATH includes /usr/local/cuda-11.1/bin
- LD_LIBRARY_PATH includes /usr/local/cuda-11.1/lib64, or, add /usr/local/cuda-11.1/lib64 to /etc/ld.so.conf and run ldconfig as root
To uninstall the CUDA Toolkit, run cuda-uninstaller in /usr/local/cuda-11.1/bin
***WARNING: Incomplete installation! This installation did not install the CUDA Driver. A driver of version at least 455.00 is required for CUDA 11.1 functionality to work.
To install the driver using this installer, run the following command, replacing <CudaInstaller> with the name of this run file:
sudo <CudaInstaller>.run --silent --driver