基于LabVIEW和MATLAB的跌倒远程监控系统

FALL REMOTE MONITORING SYSTEM BASED ON LABVIEW AND MATLAB

针对传统跌倒检测系统识别正确率较低,误报率、漏报率较高、难以满足远程实时监控的问题,设计一种基于LabVIEW和MATLAB混合编程的新型跌倒检测系统。以深度视觉传感器为数据获取源,以聚类中心点速度、高度、加速度、两中心点垂直夹角为跌倒识别特征向量,采用改进型K-means和卷积神经网络算法实现跌倒检测。实验表明,系统具有更高的识别正确率、更低的误报和漏报率、更好的鲁棒性,并满足了远程实时监控要求。

Aimed at the problems of low recognition accuracy,high false alarm rate and missing alarm rate of traditional fall detection system,and difficult to meet the remote real-time monitoring,a new fall detection system based on LabVIEW and MATLAB mixed programming was designed.The depth vision sensor was used as the data acquisition source,and the velocity,height,acceleration and the vertical angle between the two center points of the cluster center points were used as the fall recognition feature vectors.The improved k-means and convolution neural network algorithm were used to realize the fall detection.Experimental results show that the system has higher recognition accuracy,lower false positive and false negative rates,better robustness requirements,and meets the requirements of remote real-time monitoring.