Behavior extraction from multiple sensors information for human activity monitoring

摘要: A monitoring system based on multiple microelectromechanical systems (MEMS) has been developed to maintain human healthcare. Using such a MEMS system, several kinds of numerical data from types sensors can be measured. Our goal is develop intelligent with small size. In order microminiaturize the we need minimize power consumption. Therefore, have keep our simple. We propose behavior estimation method which consists SVM and fuzzy rule estimate subject's behavior. proposed two steps abstraction. First, action primitives are defined. classification trained using sample primitives. Then, classifies part into one whole expressed by sequence Next, maps actions onto defined for each second-step abstraction, as rules. From results state.