论文部分内容阅读
针对老年人跌倒检测问题,设计了一种基于加速度传感器的跌倒检测装置。装置安置于腰上采集加速度传感器数据,提取对人体运动具有较好区分度的加速度特征和躯干角度特征并对特征设定合理的阈值,建立基于多重阈值的人体跌倒检测算法,区分跌倒和人体日常生活活动。在跌倒时利用GPRS自动发出报警信息,该装置具有便携性、实时性等特点。仿真及实验表明:系统能有效识别出跌倒和日常行为,算法具有较高实时性、灵敏度和特异度。
In view of the problem of fall detection in the elderly, a fall detection device based on acceleration sensor is designed. The device is placed on the waist to collect acceleration sensor data, extract the acceleration features and torso angle features that have good discrimination of human movement and set reasonable thresholds for the features, and establish a human fall detection algorithm based on multiple thresholds to distinguish falls and human daily Living activities. GPRS automatically send alarm information when falling, the device is portable, real-time and so on. The simulation and experiment show that the system can effectively identify falls and daily behaviors, and the algorithm has high real-time, sensitivity and specificity.