Evaluation of Ice Skaters' Technical Skills
Total Page:16
File Type:pdf, Size:1020Kb
Evaluation of ice skaters’ technical skills Minkyu Kim [email protected] Motivation During a skater’s performance in an ice link, due to his or her fast and intricate movements, audiences hardly recognize how well the player’s performance is done from the perspective of technical skills. If a program can count the number of spins he or she does or detect the position and angle of foot before and after jump, which is important to judge the quality of jumps and finally give us a score for the technical skills, it will be very useful. Goal In this study, a skater’s motion is recognized through image processing in MATLAB to evaluate her technical skills. For example, to judge how well a play is done, the number of spinning a skater does is counted and the angles and relative positions of her feet when she jumps from and lands on the ground are tracked and quantified in MATLAB where a database for the judging criteria is saved beforehand. Therefore, before referees’ announcement for the final score the skater gets, we are able to calculate the score. Design A skater’s performance is recorded as a wmv file and saved in MATLAB. The skater’s motion is then read frame by frame in MATLAB. By using intensity threshold and edge detection combined with image processing techniques such as Particle Swarm Optimization (PSO), mean shift model, and so on [1-3], only her motion is highlighted frame by frame and anything else becomes dark or vice versa. By observing and analyzing each frame, the number of spinning or other technical movements is recognized. According to judging criteria referring from “2013-14 Adult Singles Free Skate Requirements [4],” her performance is evaluated and final score is shown as a result. I will not use an Android device. Preliminary Result: Captured images before and after image processing (3 sequential frames) Yuna Kim’s free skating during Vancouver Winter Olympics 2010 Images from Video Gray scale images Binary images Reference 1. “Using mean-shift tracking algorithms for real-time tracking of moving images on an autonomous vehicle testbed platform,” B. Gorry, Z. Chen, K. Hammond, A. Wallace, G. Michaelson, Proceedings of World Academy of Science, Engineering and Technology, 25, 356 (2007) 2. “Tracking and Classifying Moving Objects from Video,” Q. Zhou, J.K. Aggarwal, IEEE international workshop on performance evaluation of tracking and surveillance (PETS) (2001) 3. “Sequential Particle Swarm Optimization for Visual Tracking,” X. Zhang, W. Hu, S. Maybank, X. Li, M. Zhu, Computer Vision and Pattern Recognition, 2008. CVPR 2008. IEEE Conference on, Page: 1 – 8 (2008) 4. US Figure Skating Homepage, Technical Info, http://www.usfsa.org/content/2013- 14%20Adult%20Singles%20WBP%20Chart%20v1%2007-01-13.pdf .