Automatic front-crawl temporal phase detection using adaptive filtering of inertial signals

Area of Science:

• Sports Science
• Biomechanics
• Wearable Technology

Background:

• Accurate assessment of swimming technique is crucial for performance enhancement.
• Traditional video analysis is time-consuming and requires expert interpretation.
• Objective, real-time feedback systems are needed for efficient training.

Purpose of the Study:

• To develop and validate a novel automatic system for temporal phase detection and inter-arm coordination estimation in front-crawl swimming.
• To compare the accuracy of the proposed inertial measurement unit (IMU)-based system against a gold-standard video-based system.
• To assess the system's reliability and ease of use for swimmers and coaches.

Main Methods:

• Three waterproof IMUs (accelerometer, gyroscope) were attached to swimmers' forearms and sacrum.
• A dual-camera underwater video system served as the reference measurement.
• Swimmers performed 300m trials in varying coordination modes, analyzed using orientation estimation and adaptive change detection algorithms.

Main Results:

• The IMU system demonstrated high accuracy in estimating the index of coordination, with a mean difference of 0.2 ± 3.9% compared to video analysis.
• This accuracy was comparable to the inter-operator variability of experienced video analysts (1.1 ± 3.6%).
• The 95% limits of agreement for temporal phase estimation were within 7.9% of the cycle duration.

Conclusions:

• The proposed IMU-based system provides a valid, automatic, and user-friendly method for analyzing front-crawl swimming technique.
• This technology offers timely and objective feedback, facilitating efficient training and performance improvement.
• The system has the potential to revolutionize swimming analysis and coaching through accessible data insights.