Calf Fatigue Recognition in Heel-lift Exercise Using Video Sequences of Body Sway


BodySwayFatigue.jpg
[Abstract]
Previous analytical studies have investigated the relationship between calf fatigue and body sway measured using a force plate. However, they did not consider multiple levels of calf fatigue. Here, we propose a method for recognizing multiple levels of calf fatigue based on video sequences of body sway acquired using an overhead camera after the heel-lift exercise. For calf fatigue recognition, we extract a feature of body sway by generating a time-series signal of the head center position in the left-right and front-back directions based on medical knowledge. To evaluate the accuracy of our method, we created a dataset of 100 video sequences (20 participants x 5 calf-fatigue levels). The results show that our method can correctly recognize the calf-fatigue level with an accuracy of 40%. Furthermore, we demonstrated that for calf fatigue recognition, the accuracy of our method is superior to those of existing methods designed for human action recognition.
[Publications]



otama1.png