Logo image
An investigation into the minimum number and optimal placement of inertial measuring units for accurate golf swing detection
Thesis   Open access

An investigation into the minimum number and optimal placement of inertial measuring units for accurate golf swing detection

Divan van der Walt
M.Eng., University of Johannesburg
2025
Handle:
https://hdl.handle.net/10210/519992

Abstract

Metal oxide semiconductors, Complementary Integrated circuits -- Design and construction Metal oxide semiconductors
Golfers often face challenges in refining their swings, seeking cost-effective ways to improve their technique. Traditional coaching methods are expensive and rely heavily on the human eye, which often misses key movements due to the speed of a golf swing. To address this shortcoming, this study investigates whether inertial measurement unit (IMU) sensors can provide a more objective and data-driven method for analysing golf swing mechanics. The study investigates the minimum number of IMUs required to collect useful golf swing data, the optimal sensor placements on the body, and whether data from these locations can provide meaningful insights into a golfer’s swing mechanics. To support consistent data collection, a custom-built test rig was developed prior to the experimental studies. This was used to verify the performance of the IMUs before their use in participant trials. A two-phase experimental study was carried out. The Pilot Study involved six participants performing repeated swings with an IMU positioned at eight separate body locations, from foot to shoulder. The Main Study extended this by placing IMUs simultaneously on the most promising positions to improve data consistency and verify initial observations. Ten participants took part in the Main Study, enabling broader analysis of swing characteristics. Across both studies, accelerometer and gyroscope data were used to segment swing phases and analyse key body movements during the golf swing. The results indicated that IMUs could be used to effectively segment swing phases and provide interpretable motion data. However, not all sensor placements were equally useful; data from the hip, leading knee, and leading foot were generally less informative. The findings suggest that a minimal setup using the wrist and shoulder may be sufficient to identify swing phases and detect movement trends. Performance metrics such as peak angular velocity at these points offered quantifiable insights into swing efficiency and coordination. Additionally, multiple IMU placements enabled temporal analysis of joint sequencing, revealing subtle movement patterns consistent with biomechanical principles. Visual segmentation and the detection of posture deviations through IMU data could help players better understand and gradually improve swing consistency. Beyond their technical capabilities, IMUs are low-cost and practical, making them accessible tools for golfers and coaches outside of specialised labs. This affordability supports broader adoption in both amateur and professional coaching. Overall, the research suggests that IMUs offer a promising, objective approach to golf swing analysis, providing a foundation for further biomechanical study. Wider use of IMU technology may enable more inclusive access to quality coaching, promoting skill development and participation.
pdf
van der Walt D 201103483 - Final18.68 MBDownloadView
Open Access

Metrics

2 Record Views

Details

Logo image