Having obtained the reconstructed data in three dimensions.
We perform peak detection to extract the Four Essential
Points (FEP). In Fig. 3, the top panel shows the detected
FEP for left/right data, corresponding to the validated semantic
annotated points. For the other two subplots in this
figure, local minimum is detected. We want to draw attention
to the bottom panel of the figure. The subject shows a strong
asymmetry in his/her gait pattern. This asymmetry feature
can be developed in the future as a feature for observing
gait abnormality of Parkinson’s disease.