Event Title

User Defined Motion Counting

Session Number

F01

Advisor(s)

Mark Albert, Loyola University

Location

A-115

Start Date

28-4-2016 10:15 AM

End Date

28-4-2016 10:40 AM

Disciplines

Computer Sciences

Abstract

Already there is a large consumer market on devices that track steps and other standard measures. However, there are many more movements that can be quantified by wearable devices. This can be particularly useful to motivate a physical therapy patient’s compliance. This system was designed to be flexible enough to handle multiple movements of exercise. Wearable devices with accelerometers were attached to a participant’s body to track repetitions during exercise. Eighteen participants completed a series of 10 exercises (arm circles, bicep curls, bridges, crunches, elbow extensions, leg lifts, lunges, push ups, squats, and upper trunk rotations), twice each, for 30 seconds. Three analysis techniques were used to count the repetitions: threshold crossing, threshold crossing with a low-pass filter, and a Fourier transform with a low-pass filter. Preliminary results indicate that some exercises like arm circles and pushups are tracked well by most analysis methods, while less periodic, irregular motions like upper trunk rotations, are more difficult to track. The methods that used low-pass filtering performed reasonably well. This indicates that our system is capable of tracking a large number of activities, even ones for which it was not originally designed.


Share

COinS
 
Apr 28th, 10:15 AM Apr 28th, 10:40 AM

User Defined Motion Counting

A-115

Already there is a large consumer market on devices that track steps and other standard measures. However, there are many more movements that can be quantified by wearable devices. This can be particularly useful to motivate a physical therapy patient’s compliance. This system was designed to be flexible enough to handle multiple movements of exercise. Wearable devices with accelerometers were attached to a participant’s body to track repetitions during exercise. Eighteen participants completed a series of 10 exercises (arm circles, bicep curls, bridges, crunches, elbow extensions, leg lifts, lunges, push ups, squats, and upper trunk rotations), twice each, for 30 seconds. Three analysis techniques were used to count the repetitions: threshold crossing, threshold crossing with a low-pass filter, and a Fourier transform with a low-pass filter. Preliminary results indicate that some exercises like arm circles and pushups are tracked well by most analysis methods, while less periodic, irregular motions like upper trunk rotations, are more difficult to track. The methods that used low-pass filtering performed reasonably well. This indicates that our system is capable of tracking a large number of activities, even ones for which it was not originally designed.