Bathroom Anomaly Detection System Using Temperature Sensor Array
Hirokazu Seki · Journal of Life Support Engineering · 2019
The number of Japanese elderly people is increasing every year and the demand for life monitoring systems is also increased. Accidents in bathroom such as heat shock, slipping and falling become particularly serious in recent years. Therefore, this paper proposes an anomaly detection system in bathroom using a temperature sensor array and machine learning. The proposed system extracts the feature values of human region and detects various abnormal behaviors by One-Class SVM. In addition, the effect of improved One-Class SVM with virtual outliers is also verified. Some experiments to demonstrate the effectiveness of the proposed anomaly detection system are conducted and the results show that the high detection accuracy of at least about 90% can be realized in both the washing place and bathtub.