Viso2_ros camera info doubts

[jean821203]

In general, there is a image processing procedure in the launch, since we want rectified image as viso2 input
However, a new camera info is not published in this case after calibration.
shouldn't the calibration parameter used (focal length and principal points) in viso2 or viso2_ros be modified?
otherwise there would be flaws in projecting to 3d points....? and then the optimization...?
Im not sure if im right