基于OpenCV空巢老人跌倒检测监控系统的设计

摘要

针对人口老龄化和空巢老人问题的突出,以及健康安全问题的考虑,提出了一种基于OpenCV的空巢老人跌倒检测监控系统的设计方案.文中提出了一种全新的基于人们经验的跌倒检测方案,根据CamShift目标跟踪算法求出目标跟踪椭圆区域,以目标跟踪椭圆区域来表征老人的身体特征区域,求取出老人身体的中心点和下顶点,对这两个特征点进行加速度,角度,高度差以及停留时间的判断,来检测老人是否跌倒.实验结果表明,该方案在保证较高准确率的前提下,计算复杂度大幅度下降,能达到较好的检测效果.%In view of the problems of population aging and empty nest elderly,as well as health and safety issues,this paper puts forward a design scheme of the empty nest elderly falls detection and monitoring system based on OpenCV. This paper presents a fall detection scheme based on the experiences of the people of new,according to the CamShift target tracking algorithm for target tracking ellipse area,body feature region to characterize the old target tracking elliptic region,center point out the old body and the vertex acceleration,the two feature point height and the residence time of the judgment,to detect whether the elderly fall.The exper?imental results show that the proposed scheme can reduce the computational complexity greatly,and can achieve better detection re?sults.