Fig. 5

Root mean squared error of the stimulation based on the Kinect v2 depth camera’s COM (shown in darker color) and the pressure mat’s COP (shown in lighter color) compared against that based on the force plate’s filtered COP, separated by the type of VFBT exercise (i.e., color matching, hunting, ellipse, and bullseye shown in red, blue, green, and purple, respectively). Results are shown for the left and right plantarflexors (LPF and RPF) as well as the left and right dorsiflexors (LDF and RDF). Significant difference (p < 0.05) between sensors is indicated by *