SynchStep Blog

“

”
i think i may have found the issue with step detection. the new way of accessing the accelerometer is providing different maximums and minimums. after lots of processing, this is the cleaned up signal from the accelerometer. the first peak is the right leg, and the second hump is the left leg. my current adaptive threshold is counting BOTH legs, where my algorithm is generally set to look at only the strongest leg. going to muck with this tonight and see if i can regain the precision synchstep had in 1.1.4.

i think i may have found the issue with step detection. the new way of accessing the accelerometer is providing different maximums and minimums. after lots of processing, this is the cleaned up signal from the accelerometer. the first peak is the right leg, and the second hump is the left leg.

my current adaptive threshold is counting BOTH legs, where my algorithm is generally set to look at only the strongest leg. going to muck with this tonight and see if i can regain the precision synchstep had in 1.1.4.

More Information