1A2-P10 Anomaly Behavior Pattern Detection for Elderly People by Probabilistic Inference of Omni-Directional Image Features(Nursing and Mechatronics)

Abstract

This paper describes an anomaly behavior detection system for elderly people based on an omni-directional vision sensor. The proposed monitoring system automatically learns daily behavior patterns and detects anomaly behavior pattern using probabilistic inference approach. Bayesian Network is constructed using image feature values such as area extracted from the captured image sequence. Human trajectories are also included in image features in order to detect the prolonged motionlessness. Some experiments are performed to verify the effectiveness of the proposed system.