Howdy folks

I have logged some RGB and Depth images from a ZED camera in a rosbag, along with the camera info messages. I didn't save the point clouds as this was making the bagfiles too big. Instead, I replay my bagfiles and make use of the depthimageproc nodlets to construct the point cloud. Unfortunately I noticed that my cloud geometry is warped and I suspect that that I have been using an incorrect distortion model.

Question 1: Does the depthimageproc/pointcloudxyzrgb nodelet take the distortion model into account when constructing the point clouds from the depth image and the camera info data?

Assuming that the distortion model is used for generating the point clouds, then I would like to somehow update my bagfiles with the new distortion model. When searching for an existing tool to achieve this, I found changecamerainfo.py in bag tools :http://wiki.ros.org/bag_tools.

Question 2: I am a C++ ROS programmer, and I don't have experience with using Python in ROS. Can someone please explain to me how I would need to use changecamerainfo.py . How does the replacement YAML input file have to look like?

Thanks in advance

Galto