A Hand-Washing Support System Based on Center of Gravity in Hand Correction

Aiming at solving the problem of recognition accuracy decrease during hand-washing in existing systems due to noise produced by hand-shaking, lighting changes, and so on, we develop a novel system. The system focuses on proper hand-washing patterns to examine whether hands are being washed correctly. By comparing videos of hand-washing captured at sinks and learning model videos, the system informs users in real time whether they are using proper hand-washing patterns. Optical flows and skin-color areas are used as data features to recognize proper hand-washing patterns. Support vector machine is used as the recognition model in this system. The hand-washing pattern recognition accuracy is improved by removing noise sources and capturing pattern characteristics by labeling and using a correction that considers the center of gravity in hand.

内容要旨・審査要旨・論文本文の公開