The reanspose/inverse of rotation matrix

[NgChiho]

Dear developer:

Thank you for your work and your sharing firstly! I have tried this program in order to calculate the inverse kinematics data instantly. However I have found that the result is not so resonable until I canceled the transpose of the rotation matrix in IMUcalibrator.cpp (Line 96 and Line 142) and IMUcalibrator.h (Line 110). Then I wonder if I just got the correct result by accident while the real calculation process is not right. So I want to ask why the rotation matrix should be transposed (which means inverse in the code, I think). Thank you so much for your attention!

BTW: the IMU data I uesd is exported from Xsens

Chiho Ng

[mitkof6]

I do not know the conversation in the xsense IMUs to check the math. It is possible that the operations and assumption of what we expect q0 to be work well with NGIMUs reference frame but not with other sensors. Happy that at least you figure the order for your case.