Estimation of temporal parameters in ski jumps using wearable sensors

INTRODUCTION Take-off during ski jumping is probably the most important phase as it directly influences the whole performance [1]. It is then of particular interest to precisely identify key events such as when the take-off movement is initiated (TS), when the toes leave the ground (TO), when both legs are fully extended (LE) or when the ski jumper achieved his final flight posture (TE). In this paper, we propose a wearable system that records the accelerations and the angular velocities of ski jumpers. Based on these signals, we then automatically extract four key events during takeoff: TS, TO, LE and TE. We use these instants to derive timing information during takeoff in hill jumps.