Detecting abnormal state of elderly for service robot with H-FCM

Abstract

Detecting abnormal state of elder from sensor is very important for high-level activity inference of service robot. This paper proposes a model to solve this problem using location information. Firstly, the feature can be extracted from location information using the SLAM Map, and the vector based on l,thetas is calculated. Then approach of clustering the feature vector based on H-FCM algorithm. Experiment results demonstrate that the model is feasible to learn human state habit and can afford a judging gist to detect persons' abnormal state.