Hello,
I noticed that POSITIONAL_TRACKING_FUSION_STATUS reports VISUAL_INERTIAL when tracking with GEN_1 mode whereas just VISUAL with GEN_2 mode enabled.
How can I enable VIO while using GEN_2 positional tracking mode? setting enable_imu_fusion = True doesn’t seem to help.
In both cases I am using ZED SDK 5.0 with the same SVO2 file recorded with the X mini.
Thanks!
Hi @ZoharO,
This behavior is expected. GEN_2 is more closely integrated with the IMU compared to GEN_1, and it requires a calibration step to align the visual and inertial data. Initially, tracking is handled by the visual component while the calibration process runs in the background. Once calibration is complete, the system status transitions to VISUAL_INERTIAL.
Best regards,
Tanguy
Thank you @TanguyHardelin for clarifying this. I now notice that if left to run for long enough, it eventually does switch to VISUAL_INERTIAL. Are there any guidelines I should follow in order to expedite this transition (certain motions)? Is there any documentation available on this calibration step and on GEN_2 tracking in general? Thanks!
Hi @ZoharO,
To improve your calibration process, ensure you perform smooth movements, especially in translation. If you’re using GEN_2, be aware that it is highly sensitive to system vibrations, which might explain the extended VISUAL/IMU calibration time you’re experiencing.
We have released a new positional tracking mode called GEN_3 that addresses these issues. You can find it in the 5.0 Early Access (EA) version available here.
Best regards,
Tanguy