This forum is disabled, please visit https://forum.opencv.org
 | 1 | initial version |
For multi-view triangulation, I"m working on THIS module. You find the same point in two or more of the images and it will help you find the 3d locations, given all the camera information.
It's not perfect (the inputs need to be in a pretty specific format) but it does work.
Unlike structure from motion, this module assumes you already know the camera position and orientation, so it's fast.
Right now it only has methods for single points, and can't create a 3d model from the images, but that may come.
