Closed yangyulin closed 2 years ago
@yangyulin it sounds like this would be fixed with a trivial PR; simply changing this line right?
Also, what does the b
versus i
stand for in T_i_b
? The output convention of specifying the terms would be useful here too. e.g., something similar to T_ci: (imu0 to cam0):
for T_i_b
.
EDIT: I see that b
stands for body
, which I assume refers to the additional IMUs, while i
represents the reference imu (whose default would be imu0
)
@phil0stine, Yea. the b stands for body or based IMU.
The T_i_b represents the transformation from body frame {b} to auxiliary IMU frame {i}.
T_i_b should equals to {R_i_b, p_i_b} ( which transforms a vector from {b} to {I}).
I think you are correct. In their output results, the rotation is R_i_b (rotation from {b} to {i}), while the position is p_b_i (translation from {i} to {b}). The current results of the Kalibr are: T_i_b = {R_i_b, p_b_i}.
Hi,
I have found a bug for the multi-imu calibration.
The results generated for the T_i_b is not in the right form.
The rotation part is R_i_b while the position part is p_b_i.