Land and Underwater Estimation of Joint Angles Using MOCAP, Experimental IMUs, and OpenSense

Aquatic physical activity plays a vital role in sports training and rehabilitation, yet precise motion analysis in water remains challenging due to the limitations of traditional motion capture (MOCAP) systems. This study validates the use of experimental wearable inertial measurement units, TinyTag, combined with OpenSense software to estimate joint angles on land and underwater. The method was assessed through a two-phase approach: First, TinyTag–OpenSense joint angle estimates were validated against the gold standard motion analysis system in a controlled laboratory setting; second, the framework was applied to analyze lower-limb joint range of motion (ROM) during four hydrotherapy exercises in 15 healthy adults on land and in water. The results demonstrated good agreement with the reference system for squat and knee-to-chest (KTC) exercises, with a root mean square error below 5◦ and a Spearman correlation coefficient greater than 0.85. The underwater assessment revealed an increase of at least 10% (p < 0.005) in ROM for all exercises except squats and frontal leg swing, highlighting the supportive effects of water. This study presents a validated open source methodology for underwater motion tracking, offering a solution for clinical and research applications in aquatic rehabilitation and sports science.