How can I get acceleration ground truth from VICON?
Dear, ROS people.
I have a question about acceleration ground truth.
We have VICON system which can provide position and angles.
In order to obtain linear acceleration, double differentiation is applied to position data. Computed the acceleration, however, looks very noisy. I also tried to apply median filter but no success. Is there anyone who knows how to get nice acceleration like the linked paper?
Maybe the possible answer would be "apply low-pass filter with 4th_order Butterworth filter with 6Hz cutoff frequency".
Cheers.