Defining Groupings and Classification of Human Gait Using Correlation of Ground Reaction Force Measurements
Classification of a person’s gait through quantitative methods has wide reaching applications in security, marketing, and healthcare. This study uses ground reaction force (GRF) measurements of healthy subjects and patients with osteoarthritis (OA) to define groupings and classify subjects. In addition to grouping classes into known qualities (gender, diagnosis, etc.), this work introduces classes based on groups coming directly from correlating GRFs from different subjects. Using correlation allows new groupings to be established and more accurate classification results because continuous force time histories are compared as opposed to conventional discrete force data (i.e. peaks). Two new classes are introduced from the data which can be classified with a 92% accuracy, and the physical meaning of these classes is investigated. Comparison of a single healthy person’s walking pattern to multiple classes builds a signature that could be used to identify specific individuals. Additionally, patients suffering from OA do not correlate well with healthy groupings and can be distinguished from healthy subjects. This allows for the possibility of using GRFs to track patients’ rehabilitation. It is expected that as a patient progresses through a rehabilitation program and begins to recover, their walking patterns will become more consistent and be more highly correlated with healthy groupings.
KeywordsGait Classification Correlation Grouping GRF
Dr. Tarazaga would like to acknowledge the financial support of the John R. Jones Faculty Fellowship.
- 2.Middleton, L., Buss, A.A., Bazin, A.L., Nixon, M.S.: A floor system for gait recognition. In: Fourth IEEE Workshop on Automatic Identification Advanced Technologies, Buffalo, NY (2005)Google Scholar
- 3.Addlesee, M.D., Jones, A., Livesey, F., Samaria, F.: The ORL active floor. IEEE Pers. Commun. 35–41 (October 1997)Google Scholar
- 4.Orr, R.J., Abowd, G.D.: The smart floor: a mechanism for natural user identification and tracking. In: Proceedings of the Conference on Human Factors in Computing Systems (2000)Google Scholar