Recognition of false alarms in fall detection systems

摘要: Falls are a major cause of hospitalization and injury-related deaths among the elderly population. The detrimental effects falls, as well negative impact on health services costs, have led to great interest fall detection systems by health-care industry. most promising approaches those based wearable device that monitors movements patient, recognizes triggers an alarm. Unfortunately such techniques suffer from problem false alarms: some activities daily living erroneously reported thus reducing confidence user. This paper presents novel approach for improving accuracy which is idea identifying specific movement patterns into acceleration data. Using single accelerometer, our system can recognize these use them distinguish real falls; number alarms reduced.