You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Currently I am calibrating my 4 realsense cameras, but is facing inconsistent behavior. All the cameras are HW sync, and arranged in 4 direction facing middle. I am running on Ubuntu 16.04, Intel i7 laptop, CUDA 3.5 compute capability.
The calibration process reduce MSE, but half of the time, MSE remains high (> 1) for very long time, like 10-40 min, and sometimes, it just reduced to less than 0.2 in a minute.
I had tried several things that increase the changes of successful calibration (MSE < 0.2), 1 of it is the moving of calibration board. If the board moves too slow, MSE stucked.
I also notice the small camera pose screen with hundreds to thousands of camera poses, starts to drift to whole 3D world, expanding. Sometimes they converge back to nearing the board, most of the time never.
Any tips for higher success rate?
Thanks.
The text was updated successfully, but these errors were encountered:
Hi ARPG,
Currently I am calibrating my 4 realsense cameras, but is facing inconsistent behavior. All the cameras are HW sync, and arranged in 4 direction facing middle. I am running on Ubuntu 16.04, Intel i7 laptop, CUDA 3.5 compute capability.
The calibration process reduce MSE, but half of the time, MSE remains high (> 1) for very long time, like 10-40 min, and sometimes, it just reduced to less than 0.2 in a minute.
I had tried several things that increase the changes of successful calibration (MSE < 0.2), 1 of it is the moving of calibration board. If the board moves too slow, MSE stucked.
I also notice the small camera pose screen with hundreds to thousands of camera poses, starts to drift to whole 3D world, expanding. Sometimes they converge back to nearing the board, most of the time never.
Any tips for higher success rate?
Thanks.
The text was updated successfully, but these errors were encountered: