基于BP神经网络的摔倒动作识别方法  

Fall Recognition Method Based on BP Neural Network

作　　者：张庆宾 丁娜娜 吴海波[1] ZHANG Qingbin;DING Nana;WU Haibo(The 28th Research Institute of China Electronics Technology Group Corporation,Nanjing 210007,China;Nanjing LES Information Technology Co.Ltd.,Nanjing 210014,China)

机构地区：[1]中国电子科技集团公司第二十八研究所,南京210007 [2]南京莱斯信息技术股份有限公司,南京210014

出　　处：《指挥信息系统与技术》2021年第1期60-64,共5页Command Information System and Technology

摘　　要：高识别率的人体摔倒动作识别方法对减少老年人或病人因摔倒而导致出现危险状况具有重要意义。利用固定在人体腰部的基于姿态航向参考系统(AHRS)提供的姿态角和三轴加速度数据,提出了一种基于反向传播(BP)神经网络的摔倒动作识别方法。该方法可准确检测人体摔倒动作,以及有效区分摔倒和日常行为动作(走、跳、坐、蹲和躺等)。对不同样本的摔倒和日常行为动作试验结果表明,该方法识别率较高、稳定性较好且实用性较强。High recognition rate of human fall recognition method is of great significance to reduce the risk of the elderly patients due to fall.Based on the attitude angle and tri-axial acceleration data provided by the attitude and heading reference system(AHRS)fixed on the waist of the human body.a fall detection method based on the back propagation(BP)neural network is proposed.The method can accurately detect falls and distinguish falls from daily actions(walking,jumping,sitting,squatting and lying,etc.).The experiment results of fall and daily behavior of different samples show that this method has high recognition rate,high stability and strong practicability.

关 键 词：摔倒检测 融合加速度 BP神经网络 动作识别 

分 类 号：TP391.1[自动化与计算机技术—计算机应用技术]