Physical activity recognition using hetero-core optical fiber sensors embedded in a smart clothing
Yuya Koyama, Michiko Nishiyama, Kazuhiro Watanabe · 2018
Human activity recognition is demonstrated using smart clothing based on hetero-core fiber optics. The hetero-core optical fiber sensor monitors shape variations of clothing and detects the body motion unobtrusively for users. Motion monitor was first tested for seven activities which are standing, sitting, walking, descending stairs and climbing stairs, lying, and eating and drinking. The monitored data were processed by means of support vector machine (SVM) classifier. As a result, seven human activities were successfully recognized using hetero-core fiber optics.