Sensor-based platform-independent gait cycle detection framework
CyclePro.m - main function with the input of data file name ('fname') and sensor sampling frequency ('freq')
Correlation.m - help function to compute normalized cross-correlation
newPeak.m - help function to finalize the segmentation points for gait cycles
acc.csv - 3D accelerometer data collected from one foot in a 10-meter-walk experiment
gyro.csv - 3D gyroscope data collected from one foot in a 10-meter-walk experiment
fsr.csv - five 1D force-sensing resistor data collected under one foot in a 10-meter-walk experiment
Image files named acc-salience.jpg, acc-temp1.jpg, acc-temp2.jpg and acc-temp3.jpg provide visualization of the stepwise results in CyclePro.
If you will use this approach or the code, please cite the following paper:
@article{cyclepro, author = "Yuchao Ma and Zhila Esna Ashari and Navid Amini and Daniel Tarquinio and Kouros Nouri-Mahdavi and Mohammad Pourhomayoun and Robert D. Catena and Hassan Ghasemzadeh", title = "CyclePro: A Robust Framework for Domain-Agnostic Gait Cycle Detection", journal = "{IEEE} Sensors Journal", month = jan, year = "2019" }