Water Waste Detection and Activity Recognition at the Sink Towards Promoting Water Conservation
Shota Yamashita, Kaori Fujinami · 2020
We propose a multi-modal sensing system to detect water waste and its associated activities aimed at increasing awareness of water-conserving behavior. The sensing device adopted in this study consists of a web camera, two infrared-distance sensors, and a water-flow sensor; the web camera plays a central role by utilizing optical flow to calculate the amount of water flow from a faucet. The detection and recognition rules are constructed heuristically. An experiment showed the feasibility of the proposed system under certain lighting conditions.