Open alejandronespereira opened 3 months ago
@alejandronespereira Hi, I have the same trouble. Have you found any updates on it?
Nope, sorry. I recall I ended up publishing some tfs myself and taking those into account for a while, but as the camera is discontinued I ended up with another solution that did not use this model.
Issue Description
I am having trouble wrapping my head around the t265 IMU coordinate system. I am trying to integrate the IMU data as well as the odometry produced by the tracking camera. I see in the official documentation here that the coordinate system is a Right Hand convention with the Z pointing inwards the camera, and Y point upwards.
However, the ROS2 wrapper is outputing the IMu in a frame that does not exist on the tf that is published, camera_imu_optical_frame. i will refer to the axes denoted on this image: as XYZ (in caps) from now onwards, and the gravity vector G as pointing towards the ground.
Here are the values for the linear acceleration depending on the camera orientation:
I don''t understand why the Y axis is negative, but again the camera_imu_optical_frame has no connection to the camera that I know.
What am I missing?