Abstract: As the aging process aggravated，an active demand is recognized for the intelligent detection and rapid remote rescue when the accident falls happened on the empty nest elderly.A remote monitor and rescue community solution is developed in this paper which based on the new fall detection algorithm using the triaxial acceleration signals and the GPSOne technique.The mobile terminal-server architecture is adopted.In order to avoid leakage alarm or false alarm incidents，the terminal also supported alarm manual cancellation.Compared with the original fall detection algorithm based on the magnitude of acceleration signal vector，the new algorithm used bior3.3 wavelet and the valid detect ratio is increased from 85% to 97.75%.The rescue signal will be sent to the community center in 5 minutes through the mobile communication network along with the latitude and longitude information.Compared with vedio-based accident detection technologies，this solution has a lot of advantages such as no monitor dead angle，less calculation，no privacy leak and so on.So it has a brilliant commercial prospect in the near future.

Key words: fall monitor, triaxial acceleration, GPSOne, wavelet

摘要： 随着社会老龄化进程的加速，针对“空巢老人”意外跌倒监测和远程紧急救助的迫切需要，研发一种基于三轴向加速度的跌到检测技术，并结合GPSOne技术设计了一套用于社区的远程紧急救助的解决方案。系统采用移动无线终端-服务器模式，同时为了避免漏报警、假报警等情况发生，在终端设计中加入了人工干预技术，可以手动报警和取消报警。实验数据的统计结果表明，设计的基于bior3.3小波的三轴向传感器加速度跌倒检测算法可以将基于加速度模向量（SVM）算法的有效检测率从85%提升至97.75%，在有手机信号情况下，5分钟内可以将求救信号连同救助者的经纬度信息发送到监护中心。相比基于视频的跌倒检测技术，本方案具有监测无死角、算法简单、计算量小，不存在隐私性问题等优势，商业化前景广阔。

关键词: 跌倒监测, 三轴向加速度, GPSOne技术, 小波