A simplified approach for autonomous quality assessment of cyclic movements

To monitor performance in athletic training it is of great importance to detect and extract significant movement features. We analyzed the utility of randomly modulated periodicity and sum signal-to-modulation-noise-ratio (SMNR) for the evaluation of the foot-rest force (FRF), which is a key parameter in rowing. The foot-rest force of elite and novice athletes was recorded with an indoor rower and processed off-line. The results showed that there were significant differences between the sum SMNRs of FRF recordings obtained from elite and novice athletes. Moreover the randomly modulated periodicity feature of FRF facilitates spotting the training course as well as the efficiency of the training method especially in autonomous activity detection and rehabilitation devices. The sum SMNR feature can be used not only to distinguish expert athletes from novice athletes but also to deliver a reliable measure of performance to track improvements during the training and/or after rehabilitation.