Overview

This package contains a number of ROS 2 components, nodes, and launch files for stereo image processing.

If porting from ROS 1, please note that the stereo_image_proc node no longer exists and instead you should use stereo_image_proc.launch.py. See the Configuration page for more detail.

DisparityNode can compute disparity images from incoming stereo pairs using OpenCV’s block matching algorithm. These are best inspected using stereo_view which is available in the image_view package.

PointCloudNode can produce point clouds, which you can view in rviz, and process with PCL.

The image below shows the left/image_raw and right_image_raw. These are the raw images from each camera.

raw

Below are the left/image_rect_color and right/image_rect_color. These are the rectified images from each camera. The red lines show that the same point in the real world lies on the same horizontal line in each rectified image.

rectified

The resulting disparity image is shown below, viewed with stereo_view from the image_view package.

disparity

Indices and tables