Replies: 2 comments 1 reply
-
Hi there, yes, the
|
Beta Was this translation helpful? Give feedback.
-
I see. So if I understood correctly, colmap is responsible for generating every camera pose. And the unit cube is aligned with the pose of the first camera, assuming it has a 45º (approx) vertical angle? I guess what I'm trying to figure out is how to generate datasets in a way that the unit cube is better aligned with the floor of the scene, so that I can crop the aabb cube and generate more relevant meshes without all the noise from unseen areas. |
Beta Was this translation helpful? Give feedback.
-
I'm trying to rotate initial camera pose so that the scene aligns better with the unit cube. I've tried modifying the
camera_angle_x
andcamera_angle_y
but I see no change at all, so I guess each picture'stransform_matrix
is fully responsible for the camera pose? Is there another way to do this? My hopes is that by aligning the scene better, it will make marchin cube meshes easier to handleBeta Was this translation helpful? Give feedback.
All reactions