Skip to Main Content
 

Global Search Box

 
 
 
 

Files

File List


PidaparthyH.the (final comments 1).pdf (2.2 MB)

ETD Abstract Container

Abstract Header

Recognizing and Detecting Errors in Exercises using Kinect Skeleton Data

Pidaparthy, Hemanth
http://rave.ohiolink.edu/etdc/view?acc_num=akron1430912344

Abstract Details

2015, Master of Science, University of Akron, Computer Engineering.
A novel approach to recognizing and correcting errors in exercise activity based on skeletal joint data obtained from the Kinect 2.0 sensor is presented. Many approaches in the literature for analyzing human motion focus on training a classifier to recognize and/or rank the motions. While effective in some situations, the computational costs of training the models, the unavailability of reference motions and the inability to provide real-time guidance and feedback limit the utility of such approaches for empowering wellness management. A classification technique is used to recognize exercises and a geometric characterization of poses is used to detect errors in the recognized exercises. All the features used are extracted from the time-series data collected from a Microsoft Kinect 2.0 camera. Expert domain knowledge was easily integrated to identify errors in exercise performance. The simplicity and the low computational costs, make this approach useful for providing real-time feedback and guidance to participants, thus improving exercise adherence. Experimental results that demonstrate the viability of the approach are presented. In the future, this approach can be extended to a wider range of exercises and similar techniques can be applied to address related problems in rehabilitation, surveillance and remote user interaction.
Shivakumar Sastry, Dr. (Advisor)
Forrest Bao, Dr. (Committee Member)
Jin Kocsis, Dr. (Committee Member)
Victor Pinheiro, Dr. (Committee Member)
53 p.
Computer Engineering
Microsoft Kinect Sensor, Personalized Wellness Management, Exercise Data Analytics

Recommended Citations

Citations

  • Pidaparthy, H. (2015). Recognizing and Detecting Errors in Exercises using Kinect Skeleton Data [Master's thesis, University of Akron]. OhioLINK Electronic Theses and Dissertations Center. http://rave.ohiolink.edu/etdc/view?acc_num=akron1430912344

    APA Style (7th edition)

  • Pidaparthy, Hemanth. Recognizing and Detecting Errors in Exercises using Kinect Skeleton Data. 2015. University of Akron, Master's thesis. OhioLINK Electronic Theses and Dissertations Center, http://rave.ohiolink.edu/etdc/view?acc_num=akron1430912344.

    MLA Style (8th edition)

  • Pidaparthy, Hemanth. "Recognizing and Detecting Errors in Exercises using Kinect Skeleton Data." Master's thesis, University of Akron, 2015. http://rave.ohiolink.edu/etdc/view?acc_num=akron1430912344

    Chicago Manual of Style (17th edition)

akron1430912344
545
© 2015, all rights reserved.
This open access ETD is published by University of Akron and OhioLINK.