cv_bridge converts between ROS 2 image messages and OpenCV image representation for perception applications. As follows:

This ros2 branch initially derives from porting on the basis of ros kinetic branch

Firstly, it assumes that the ROS 2 core has already been installed, please refer to ROS 2 installation to get started.

OpenCV3+ is a must to install, please refer to the official installation guide from OpenCV Tutorials Since ROS 2 uses Python 3, please make sure that python3-numpy is installed, or install like this:

sudo apt install python3-numpy

The cv_bridge python backend still has a dependency on python boost (equal or higher than 1.58.0), and install them as follows in Ubuntu:

sudo apt install libboost-python-dev

cd <YOUR_ROS2_WORKSPACE>/src
git clone https://github.com/ros-perception/vision_opencv.git -b ros2
cd ..
colcon build --symlink-install

Python tests have a dependency on OpenCV Python support. To install it:

sudo apt install python3-opencv

Next to prepare runtime environment and run tests:

source <YOUR_ROS2_WORKSPACE>/install/local_setup.bash
colcon test

• boost endian: remove boost endian APIs with standard C++ 11 or higher instead
• Not tested with Windows or macOS environments so there may be issues building or running