High Speed Motion Tracking for Weightlifting based on Correlation Coefficient Template Matching
Gihan Kuruppu1, S. R. Kodituwakku2, U. A. J. Pinidiyaarachchi3
1Gihan Kuruppu, Postgraduate institute of Science, University of Peradeniya Sri Lanka.
2S. R. Kodituwakku, Department of Statistics and Computer Science,University of Peradeniya, Sri Lanka.
3U. A. J. Pinidiyaarachchi, Department of Statistics and Computer Science,University of Peradeniya, Sri Lanka.
Manuscript received on January 01, 2013. | Revised Manuscript received on January 02, 2013. | Manuscript published on January 05, 2013. | PP: 439-442 | Volume-2, Issue-6, January 2013. | Retrieval Number: F1221112612/2013©BEIESP
Open Access | Ethics and Policies | Cite
© The Authors. Published By: Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: Tracking of motion in weightlifting is a challenging task. This paper presents a method for tracking Weightlifting bar movement of an athlete. The proposed method uses three different template creation techniques for tracking and applies Wilcoxon Signed rank test to evaluate the results. The method was tested with normally captured data (30fps) and data captured in slow motion (100fps) separately. For the normally captured data, combined dynamic templates method has the highest accuracy of 95% in tracking the trajectory. For slow motion data Wilcoxon Signed Rank generated equal results for all three template creation methods.
Keywords: Tracking Algorithms, Block Matching Algorithms, Fast Motion Tracking