Development Of Multi-mode Information System For Elderly Fall Detection

With the continuous development of China's aging population,the proportion of aging population is increasing.Elderly people generally suffer from degeneration of body function and prone to fall.At present,most of the detection equipment for the daily attitude of falls,syncope and other critical falls do not have a reliable and effective detection method,for the fall after the degree of emergency does not have the corresponding alarm level.Aiming at the increasing demand of fall detection and alarm for the elderly,especially for the cases of syncope or shock fall of the elderly living alone,a fall detection device based on multi-modal sensor information fusion is designed.This thesis focuses on the research and development of multi-modal information fusion based on the fall detection software and hardware equipment,through the introduction of new detection and alarm methods,the critical fall of reliable detection and alarm.The main contents of this thesis include the following aspects:1.A method of critical fall detection and alarm based on physiological sensor and fall score scale is proposed.The fusion of physiological sensor can effectively evaluate the occurrence of critical fall.By introducing the result output of fall score scale,the alarm level can be output effectively,and the result will be self-optimized learning.2.A cascade algorithm for fall detection based on multi-modal fusion is proposed,which integrates the first-line clinical investigation and medical scale.It is considered that the fall occurs mainly in the case of body displacement,and the heart rate rises after the fall because of the stress reaction.Therefore,the physical posture of the fall is judged first,and then the heart rate is detected.The cascade detection method improves the accuracy of fall detection.3.Development of a fall detection module based on multimodal information fusion: The low-power Bluetooth SOC chip is used as communication and main controller,and the detection circuit of multi-MEMS sensors including three-axis acceleration sensor,height sensor and heart rate sensor is integrated.The whole system is wearable,low-power and low-cost.Through the simulated fall detection and analysis of the experimental objects,the expected targets were achieved.To sum up,this thesis studies and develops the fall detection equipment based on multi-modal sensor information fusion and the corresponding data fusion algorithm,which has the advantages of small size,low power consumption,light weight and high accuracy.It has a high application value in the fall detection of the elderly,especially in the rehabilitation of the elderly after surgery.