The step counting is based on minimal peak distance, minimal peak prominence, dynamic thresholding and vibration elimination in combination with peak detection and periodicity, similarity features. It solves over-counting, under-counting and false walking problems and requires a low computation cost.
We have been published 4 files on github.com, it includes:
- 500steps_1
- Our_Proposed_Method_500_steps_detected_over_500_steps_execution_1.p
- 500steps_2
- Our_Proposed_Method_500_steps_detected_over_500_steps_execution_2.p
where, 500steps_1 and 500steps_2 are two recorded data by volunteers in free walking with 500 steps execution
Our_Proposed_Method_500_steps_detected_over_500_steps_execution_1.p and Our_Proposed_Method_500_steps_detected_over_500_steps_execution_2.p are step counting results by our proposed method on 500steps_1 and 500steps_2 data respectively.