Skip to content

Combined pointcloud transforms to prevent rounding errors #1182

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
jPaek1002 wants to merge 1 commit into
base: master
Choose a base branch
from
Open

Combined pointcloud transforms to prevent rounding errors #1182

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
15 changes: 11 additions & 4 deletions python-sdk/nuscenes/nuscenes.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -889,13 +889,20 @@ def map_pointcloud_to_image(self,

# Second step: transform from ego to the global frame.
poserecord = self.nusc.get('ego_pose', pointsensor['ego_pose_token'])
pc.rotate(Quaternion(poserecord['rotation']).rotation_matrix)
pc.translate(np.array(poserecord['translation']))
sensor_timestamp_ego_pose = np.eye(4)
sensor_timestamp_ego_pose[:3, :3] = Quaternion(poserecord['rotation']).rotation_matrix
sensor_timestamp_ego_pose[:3, 3] = np.array(poserecord['translation'])

# Third step: transform from global into the ego vehicle frame for the timestamp of the image.
poserecord = self.nusc.get('ego_pose', cam['ego_pose_token'])
pc.translate(-np.array(poserecord['translation']))
pc.rotate(Quaternion(poserecord['rotation']).rotation_matrix.T)
image_timestamp_ego_pose = np.eye(4)
image_timestamp_ego_pose[:3, :3] = Quaternion(poserecord['rotation']).rotation_matrix
image_timestamp_ego_pose[:3, 3] = np.array(poserecord['translation'])

# Compute the transformation matrix from point sensor timestamp to image timestamp.
sensor_to_image_matrix = np.linalg.inv(image_timestamp_ego_pose).dot(sensor_timestamp_ego_pose)
pc.rotate(sensor_to_image_matrix[:3, :3])
pc.translate(sensor_to_image_matrix[:3, 3])

# Fourth step: transform from ego into the camera.
cs_record = self.nusc.get('calibrated_sensor', cam['calibrated_sensor_token'])
Expand Down