about rotation order

[yu8ri1]

I would like to know the rotation order of the rotaion at the Euler angles of the front camera and lidar in this image.

[whyekit-motional]

@yu8ri1 pls clarify what you mean by "rotation order"

[yu8ri1]

I am trying to express the rotation angle (roll->pitch->yaw) in a simulator with different coordinate systems in the coordinate system of each sensor in nuscenes. I think it is possible to convert the coordinate system by switching the axes and changing the sign in the roll, pitch, and yaw states. In this case, I would like to match the order of rotation when nuscense is expressed as roll->pitch->yaw. What I want to ask is the rotation order when nuscense is expressed as roll->pitch->yaw.

[whyekit-motional]

@yu8ri1 the transformation is simply done as a matrix operation (i.e. there is no order of roll / pitch / yaw) - pls see here for an example: https://github.com/nutonomy/nuscenes-devkit/blob/9b165b1018a64623b65c17b64f3c9dd746040f36/python-sdk/nuscenes/nuscenes.py#L884

[yu8ri1]

Thanks for the reply . I understand about the order of rotation. The question is about rotation and translation for each sensor." .json" file are the rotations and translations of each sensor adapted to a specific coordinate system?

[whyekit-motional]

@yu8ri1 the rotation and translation of each sensor is with respect to the ego frame (pls see https://github.com/nutonomy/nuscenes-devkit/blob/master/docs/schema_nuscenes.md#calibrated_sensor for more details)

[LeningWang]

Thanks for the reply . I understand about the order of rotation. The question is about rotation and translation for each sensor." .json" file are the rotations and translations of each sensor adapted to a specific coordinate system?

Maybe we have same problem. You can refer to https://github.com/nutonomy/nuscenes-devkit/issues/1001 for the discussion