CCTag has a CPU and a GPU implementation. The GPU implementation requires an NVIDIA GPU card with a CUDA compute capability >= 3.5. You can check your NVIDIA GPU card CC support here or on the NVIDIA dev page. If you do not have a NVIDIA card you will still able to compile and use the CPU version.

Here are the minimum hardware requirements for CCTag:

Minimum requirements

Operating systems

Windows x64, Linux, macOS


Recent Intel or AMD cpus

RAM Memory

8 GB

Hard Drive

No particular requirements


NVIDIA CUDA-enabled GPU (compute capability >= 3.5)


CCTag depends on the following libraries:

  • Eigen3 >= 3.3.4

  • Boost >= 1.66

  • OpenCV >= 3.1

  • TBB >= 4.0


In order to have Cuda support on Windows, at least Eigen 3.3.9 is required

CCTag can be installed from the following package managers.


vcpkg is a cross-platform (Windows, Linux and MacOS), open-source package manager created by Microsoft.

Since v1.0.0 of the library it is possible to build and install the library through vcpkg on Linux, Windows and MacOS by running:

vcpkg install cctag[cuda,apps]

where cuda and apps are the options to build the library with the cuda support and the sample applications, respectively.


conan is a decentralized and multi-platform package manager.

Since v1.0.1, you can install CCTag from conan by running:

conan install cctag/1.0.1@

where 1.0.1@ is the version you want to install. See CCTag as third party for how to use CCTag as third party in your project.

Building the library

Building tools

Required tools:

  • CMake >= 3.14 to build the code

  • Git

  • C/C++ compiler supporting the C++14 standard (gcc >= 5, clang >= 3.4, msvc >= 2017)

Optional tool:

  • CUDA >= 9.0


On Windows, there are compatibility issues to build the GPU part due to conflicts between msvc/nvcc/thrust/eigen/boost.



vcpkg can be used to install all the dependencies on all the supported platforms. This is particularly useful on Windows. To install the dependencies:

vcpkg install

You can add the flag --triplet to specify the architecture and the version you want to build. For example:

  • --triplet x64-windows will build the dynamic version for Windows 64 bit

  • --triplet x64-windows-static will build the static version for Windows 64 bit

  • --triplet x64-linux-dynamic will build the dynamic version for Linux 64 bit

and so on. More information can be found here


On Linux you can install from the package manager:

For Ubuntu/Debian package system:

sudo apt-get install g++ git-all libpng12-dev libjpeg-dev libeigen3-dev libboost-all-dev libtbb-dev

For CentOS package system:

sudo yum install gcc-c++ git libpng-devel libjpeg-turbo-devel eigen3-devel boost-devel      tbb-devel


On MacOs using Homebrew install the following packages:

brew install git libpng libjpeg eigen boost tbb

Getting the sources

git clone

CMake configuration

From CCTag root folder you can run cmake:

mkdir build && cd build
cmake ..
make -j `nproc`

On Windows add -G "Visual Studio 16 2019" -A x64 to generate the Visual Studio solution according to your VS version (see CMake documentation).

If you are using the dependencies built with VCPKG you need to pass -DCMAKE_TOOLCHAIN_FILE=path/to/vcpkg/scripts/buildsystems/vcpkg.cmake at cmake step to let it know where to find the dependencies.

Otherwise you can specify the path where each dependency can be found (if not installed in system folders) by passing its related path. For example, for OpenCV you can pass -DOpenCV_DIR=path/to/opencv/install/share/OpenCV/ to tell where the OpenCVConfig.cmake file can be found.

CMake options

CMake configuration can be controlled by changing the values of the following variables (here with their default value)

  • CCTAG_WITH_CUDA:BOOL=ON to enable/disable the Cuda implementation

  • BUILD_SHARED_LIBS:BOOL=ON to enable/disable the building shared libraries

  • CCTAG_ENABLE_SIMD_AVX2:BOOL=OFF to enable/disable the AVX2 optimizations

  • CCTAG_BUILD_TESTS:BOOL=OFF to enable/disable the building of the unit tests

  • CCTAG_BUILD_APPS:BOOL=ON to enable/disable the building of applications

  • CCTAG_BUILD_DOC:BOOL=OFF to enable/disable building this documentation

So if you do not want to build the Cuda part, you have to pass -DCCTAG_WITH_CUDA:BOOL=OFF and so on.

CCTag as third party

When you install CCTag a file CCTagConfig.cmake is installed in <install_prefix>/lib/cmake/CCTag/ that allows you to import the library in your CMake project. In your CMakeLists.txt file you can add the dependency in this way:

 1# Find the package from the CCTagConfig.cmake
 2# in <prefix>/lib/cmake/CCTag/. Under the namespace CCTag::
 3# it exposes the target CCTag that allows you to compile
 4# and link with the library
 5find_package(CCTag CONFIG REQUIRED)
 7# suppose you want to try it out in a executable
 8add_executable(cctagtest yourfile.cpp)
 9# add link to the library
10target_link_libraries(cctagtest PUBLIC CCTag::CCTag)

Then, in order to build just pass the location of CCTagConfig.cmake from the cmake command line:

cmake .. -DCCTag_DIR=$CCTAG_INSTALL/lib/cmake/CCTag/

If you are using conan for your project then you need to add cctag to your conanfile.txt:



and when building you may need to follow these steps:

mkdir build
cd build
conan install .. -s build_type=Release
cmake .. -DCMAKE_TOOLCHAIN_FILE=conan_toolchain.cmake -DCMAKE_BUILD_TYPE=Release
cmake --build . --config Release

Docker image

A docker image can be built using the Ubuntu based Dockerfile, which is based on nvidia/cuda image ( )

Building the dependency image

We provide a Dockerfile_deps containing a cuda image with all the necessary CCTag dependencies installed.

A parameter CUDA_TAG can be passed when building the image to select the cuda version. Similarly, OS_TAG can be passed to select the Ubuntu version. By default, CUDA_TAG=10.2 and OS_TAG=18.04

For example to create the dependency image based on ubuntu 18.04 with cuda 8.0 for development, use

docker build --build-arg CUDA_TAG=8.0 --tag alicevision/cctag-deps:cuda8.0-ubuntu18.04 -f Dockerfile_deps .

The complete list of available tags can be found on the nvidia dockerhub page

Building the CCTag image

Once you built the dependency image, you can build the cctag image in the same manner using Dockerfile:

docker build --tag alicevision/cctag:cuda8.0-ubuntu18.04 .

Running the CCTag image

In order to run the image nvidia docker is needed: see the installation instruction. Once installed, the docker can be run, e.g., in interactive mode with

docker run -it --runtime=nvidia alicevision/cctag:cuda8.0-ubuntu18.04

Official images on DockeHub

Check the docker hub CCTag repository for the available images.