ZED 2i Positioning tracking

Hello, I have a question about ZED 2i with positioning tracking api.


The document states that computer vision and stereo SLAM technology was used to estimate the position.

Do you mean that only vision technology was used for ZED 2i position estimation, or that an IMU was integrated through sensor fusion such as a Kalman filter?

Thank you!

Hi @LeeHyokBeen
the positional tracking uses VIO algorithms, fusing visual odometry with inertial information.
Kalman Filters are obviously used, but I cannot provide more details because this is a piece of closed information.

Thanks for the answer!

I have another question while using ZED 2i with ZED_positional_Tracking c++ API.

I can get position data (x, y, z) fused with IMU and SLAM, as shown below.


Is there a calibrated (Visual Inertial Odometry) translation velocity parameter (vx,vy,vz) like the calibrated position data(VIO) or do I need to divide the position by the time?

Thank you!

We provide also velocity information. All the available information is described in the documentation of the sl::Pose C++ class.

Thanks for the answer!

I was trying to find the information of linear velocity (vx, vy, vz).
However, I can’t find any function or parameter for linear velocity (speed of translation).

I can only find angular velocity or linear acceleration.

How can I reach the calibrated linear velocity of the ZED 2i using the C++ API?

Thank you

Hi @LeeHyokBeen,

I think you can access the velocity from the twist values in Pose.
You could also derive it from the IMUData acceleration and timestamp.