中文摘要：

针对老年人摔倒时可能因无法得到及时救助而发生危险的情况,设计了基于WSN的人体姿态辨识系统,包括人体姿态感知节点、数据采集节点和实时监控管理平台等三部分,能对多目标人体姿态进行实时辨识,一旦人体姿态存在异常能立刻进行报警,从而保证了目标的人身安全.实验检测结果表明,系统能在不影响使用者日常生活的前提下,同时对多个目标的人体姿态进行正确辨识,当监测到目标处于跌倒姿态时,可以及时进行报警,满足了对老年人活动状态进行实时监控的需求,

英文摘要：

For the case may be due to inability to get rescue in time when the elders fall down, this paper proposed a human posture recognition system based on WSN, which consists of the body posture recognition nodes, the data collection nodes and the real-time monitoring platform. The system can recognize real-time multi-target human gesture and send alarm as soon as the human posture is abnormal in order to ensure the personal safety. The test results showed that the system can recognize multiple targets of human posture correctly based on no affecting on daily life. When the target is in the fallen-down posture, the system can monitor and alarm timely. It will meet the requirement of active monitoring for elders.