UTS Capstone utilizing machine learning to classify gait pattern

Construct IMU Platform (to test and place on body = IMU + battery) Three Calibration Modes:

• Sitting/Standing
• Walking
• Running

Implement using LabView wirelessly via Bluetooth Using K means clustering

• Organise data plots
• • Each individual gait test e.g. walking

Perform test in an environment of constant temperature No external factors such as: wind, air con, heater

• Since utilizing 3-axis accelerometers only, will require that sensor to be placed where the swing occurs during gait as walking and running have a very distinct acceleration during the gait.
• Ankle
• Base of foot #1

Use auto-calibration of output of IMU in LabVIEW

• K-means clustering to filter outputs
• • If/For statements to determine which movement is occurring i.e. walking, running or standing.
• • Each mode has a separate screen to display activity OR
• • One screen and multiple plots (Walking, Running, Standing

• Set a threshold value i.e. a certain magnitude of acceleration specific to walking (repeat for running)
• Code that all other samples to be Stationary/Standing/Sitting

• How to read IMU data from LabVIEW
• How to export data to LabVIEW