search

hailanyi / 3D-Multi-Object-Tracker

A project for 3D multi-object tracking
Apache License 2.0
302 stars 55 forks source link

The pose file is obtained from the oxts files provided by KITTI tracking dataset using the official matlab tools. #17

Closed Sonne-Zhu closed 1 year ago

Sonne-Zhu commented 1 year ago 
    The pose file is obtained from the oxts files provided by KITTI tracking dataset using the official matlab tools.

Originally posted by @hailanyi in https://github.com/hailanyi/3D-Multi-Object-Tracker/issues/11#issuecomment-1279639990

Thank you for your great work！I have a question about pose.txt you provided:

Using official matlab tools directly only seems to get the pose in IMU/GPS coordinate system, so did you convert it to lidar coordinate system? Because I found that your code seems to use only the external parameter between lidar and camera, i.e. velo_to_cam, but not use external parameter between lidar and imu.

So is the pose.txt you provided in the lidar coordinate system or in the IMU/GPS coordinate system?

hoangduyloc commented 1 year ago 
    The pose file is obtained from the oxts files provided by KITTI tracking dataset using the official matlab tools.

Originally posted by @hailanyi in #11 (comment)

Thank you for your great work！I have a question about pose.txt you provided:

Using official matlab tools directly only seems to get the pose in IMU/GPS coordinate system, so did you convert it to lidar coordinate system? Because I found that your code seems to use only the external parameter between lidar and camera, i.e. velo_to_cam, but not use external parameter between lidar and imu.

So is the pose.txt you provided in the lidar coordinate system or in the IMU/GPS coordinate system?

Hi, I think this might help you: https://github.com/xinshuoweng/AB3DMOT/blob/master/AB3DMOT_libs/kitti_oxts.py

I also wonder the case, if we dont have the pose, we have custom LIDAR only dataset, and after prediction we have 3D bboxes only. How can we apply this work?

Sonne-Zhu commented 1 year ago 
    The pose file is obtained from the oxts files provided by KITTI tracking dataset using the official matlab tools.

Originally posted by @hailanyi in #11 (comment) Thank you for your great work！I have a question about pose.txt you provided: Using official matlab tools directly only seems to get the pose in IMU/GPS coordinate system, so did you convert it to lidar coordinate system? Because I found that your code seems to use only the external parameter between lidar and camera, i.e. velo_to_cam, but not use external parameter between lidar and imu. So is the pose.txt you provided in the lidar coordinate system or in the IMU/GPS coordinate system?

Hi, I think this might help you: https://github.com/xinshuoweng/AB3DMOT/blob/master/AB3DMOT_libs/kitti_oxts.py

I also wonder the case, if we dont have the pose, we have custom LIDAR only dataset, and after prediction we have 3D bboxes only. How can we apply this work?

Thanks for the reply, so it looks like pose.txt is in imu/GPS coordinate, wouldn't it be necessary to convert it to lidar coordinate? For your question, I think if there is no GPS to use as ground truth, performing an open source high precision lidar slam algorithm, such as LeGO-LOAM, to get lidar odometry is a viable method.

hailanyi commented 1 year ago

The pose is optional. If you don't have a pose, you can enter an 4x4 identity matrix instead.