I'm analyzing gyroscope + accelerometer data acquired from an IMU for a run of a water skiing slalom course, sampled at 100Hz.
The double spikes represent spurious ~15G jitter from crossing both wakes.
My question is two-fold:
How can I use the fact that the skiier traverses through known positions along the course to determine a dead-reckoning for spatial position?
How do I filter/account for the two impulses that don't represent the true acceleration of the skiier?