Description
I imported a set of camera views into MeshLab (in the bundler.out format) following performing 3D reconstruction to obtain a sparse point cloud. That worked and the point cloud was at the 'centre' with camera poses from different directions looking at it. I then brought this into MeshLab via the aforementioned bundler.out format. However, following scaling the cameras (they are by default hundreds of time larger to begin with), it looked like everything was fine and I could see the point cloud at the centre with camera poses around it just like in the tool from which I exported it. However, when I attempt to look through each camera, only the photo associated with that view is visible and there is no point cloud. There is no problem in terms of me not knowing how to make the image more and more transparent in order to see the 3D point cloud. The point cloud just disappears when getting into the camera view but is there when switching out of it and 'zooming out'.
This import and viewing works in some cases and fails in other cases even though the process I followed is the same. Is this a known issue? Has anyone got past this?