Conference paper
A Flexible method of jump and high intensity event detection
Proceedings of the 14th Australasian Conference on Mathematics and Computers in Sport, pp.101-106
Australasian Conference on Mathematics and Computers in Sport (ANZIAM Mathsport 2018), 14th (Sunshine Coast, Australia, 25-Jul-2018 - 28-Jul-2018)
ANZIAM MathSport
2018
Abstract
With success in competitive sports often relying on players making quick, explosive movements High Intensity Events (HIE's) have become fundamental in assessing athlete performance. Their detection has traditionally relied on video analysis but with many athletes and elite teams now wearing inertial sensors to record their movement there is a demand to develop classification algorithms that find a balance in accuracy between specificity of the athlete, and robustness of detection over a range of athletes or HIEs. In this research we develop a novel approach to detect jumps in a complex sporting environment and apply this method to other sporting applications. Using elite netball as a test case acceleration data from a training session was analysed and key parameters of the event were defined. Code was developed using a single acceleration axis to automatically detect jumps as well as identify the time it occurred and several performance indicators including flight time and maximum acceleration at take-off and landing. Applying the code to an elite netball match it detected 100% of jumps identified by video analysis with no false positives. Time of take-off and landing were within 0.04sec being 1 frame of video taken at 25 frames/sec. The unchanged code was then tested in several sports and successfully classified other HIEs in athletics, and water sports. The specificity and robustness of this method will have wide practical applications in the detection of jumps and the potential to identify other HIEs in many sports. This will enable coaches and sport scientists to classify these events with no pre-conditioning of the data, minimal change to the model, and have the potential to produce results in real time. This method also shows great potential in the wider application of human movement detection in the general fitness and health sectors.
Details
- Title
- A Flexible method of jump and high intensity event detection
- Authors
- Paul Smith (Author) - University of the Sunshine Coast, Queensland, Faculty of Science, Health, Education and EngineeringAnthony Bedford (Author) - University of the Sunshine Coast - Faculty of Science, Health, Education and Engineering
- Contributors
- Ray Stefani (Editor)Anthony Bedford (Editor)
- Publication details
- Proceedings of the 14th Australasian Conference on Mathematics and Computers in Sport, pp.101-106
- Conference details
- Australasian Conference on Mathematics and Computers in Sport (ANZIAM Mathsport 2018), 14th (Sunshine Coast, Australia, 25-Jul-2018 - 28-Jul-2018)
- Publisher
- ANZIAM MathSport
- Organisation Unit
- School of Health and Sport Sciences - Legacy; School of Science, Technology and Engineering; School of Health and Behavioural Sciences - Legacy; University of the Sunshine Coast, Queensland; Faculty of Science, Health, Education and Engineering
- Language
- English
- Record Identifier
- 99450785402621
- Output Type
- Conference paper
Metrics
460 Record Views