Questions about the coordinate axis

[aiming1998]

@mmattar @xpchuan-95 @malcolmgreaves @rickyrauch @brwillia @nisseknudsen
Hi~
I read the paper
[https://arxiv.org/pdf/2112.12610.pdf ]
I think it is a mistake about Fig 1.(a) Sensor placement
In this figure, the car faces to the right, but the positive direction of the Y-axis of the lidar faces back
Where is the origin of the world coordinate axis you mentioned? Where is it located on the vehicle?
The point cloud in frame 0 takes this coordinate axis as the origin, but the values of Y-axis and x-axis of lidar are positive, which is impossible for panderGT
I think the x-axis direction of solid-state lidar should be both positive and negative
Can you give the correct definition of the coordinate system?
I think the point clouds of these two lidars are calibrated to a unified coordinate system. Where is the origin of this coordinate system?
Thanks~

[Leozyc-waseda]

Hi,
I am also very confused about this problem, but I successfully converted the coordinates of Pandaset to the standard coordinates of openpcdet using this ego coordinate system, which works well.

https://github.com/open-mmlab/OpenPCDet/blob/e1bfcecbf4966fe417bf56fa60d6e177a1909724/pcdet/datasets/pandaset/pandaset_dataset.py#L171-L180

hope this helps.

[aiming1998]

Hi, I am also very confused about this problem, but I successfully converted the coordinates of Pandaset to the standard coordinates of openpcdet using this ego coordinate system, which works well.

https://github.com/open-mmlab/OpenPCDet/blob/e1bfcecbf4966fe417bf56fa60d6e177a1909724/pcdet/datasets/pandaset/pandaset_dataset.py#L171-L180

hope this helps.

Thank you a thousand times~