Label dataset with CloudCompare about crlfnet HOT 6 CLOSED

Message that will be displayed on users first issue

from crlfnet.

Absolutely yes, see link the official custom dataset support was initialized by this repo.

from crlfnet.

loc_lidar = np.concatenate([np.array((float(loc_obj[2]), float(-loc_obj[0]), float(loc_obj[1]-2.3)), dtype=np.float32).reshape(1,3) for loc_obj in loc])

For the above line of code, it is the kitti format obtained by labeling with labelcloud. I get the geometric center directly. Can I replace this line of code with the following?

loc_lidar = np.concatenate([np.array((float(loc_obj[0]), float(loc_obj[1]), float(loc_obj[2])), dtype=np.float32).reshape(1,3) for loc_obj in loc])

from crlfnet.

No problem, BTW you should make sure that the coordinates (loc_obj[0], loc_obj[1], loc_obj[2]) are aligned with the world coordinates system (the same as the point coordinates in), note that the sequences are changed in ((float(loc_obj[2]), float(-loc_obj[0]), float(loc_obj[1]-2.3))

from crlfnet.

No problem, BTW you should make sure that the coordinates (loc_obj[0], loc_obj[1], loc_obj[2]) are aligned with the world coordinates system (the same as the point coordinates in), note that the sequences are changed in ((float(loc_obj[2]), float(-loc_obj[0]), float(loc_obj[1]-2.3))

How do I see if my point cloud is aligned with the world coordinate system?

from crlfnet.

@Lizhinwafu Emm... You could learn relative knowledge first.

from crlfnet.