Currently the "groupfind" function used with in HkLee, takes non-zero sequences with 2 or more values, Those sequences indicate a contact. However 2 samples in a signal of 120Hz might be noise and not referring to an actual contact. Future improved versions could adjust the groupfind function to include sequences equal or longer that the minimum possible contact time.
Further, we also consider "open" interevals at the start and the end of the sequence, which can lead to addtional detected ICs, eventhough we only detected a partial "signal-shape" representing a stride
Currently the "groupfind" function used with in HkLee, takes non-zero sequences with 2 or more values, Those sequences indicate a contact. However 2 samples in a signal of 120Hz might be noise and not referring to an actual contact. Future improved versions could adjust the groupfind function to include sequences equal or longer that the minimum possible contact time.
Further, we also consider "open" interevals at the start and the end of the sequence, which can lead to addtional detected ICs, eventhough we only detected a partial "signal-shape" representing a stride