基于OpenMv的跌倒检测算法设计与实现

Design and Implementation of Drop Detection Algorithm Based on OpenMv

摘要: 现如今，社会上老年人的数量正在持续增长，而伴随着的也是日益增多的老年人的安全问题，例如跌倒、突发疾病、失智等一系列对老年人生命产生威胁的问题，本文就其中的跌倒问题提出了一种基于OpenMv的跌倒检测算法，主要通过对OpenMv摄像头捕捉到的画面进行图像运算得到老年人的跌倒状态。本文提出的跌倒检测主要通过OpenMv自带的函数库以及三帧差分算法来实现，然后通过调整老年人身体中心坐标的波动范围来判断老年人是否已经跌倒。

Abstract: Today, the number of elderly people in society is on the rise, but also accompanied by a growing number of elderly security problems, such as falls, sudden illness, dementia, and a series of a threat to the elderly life problems, and in this paper, the fall problem is proposed based on a OpenMv fall detection algorithm, mainly through OpenMv cameras to capture images for image arithmetic for the elderly fall state. The fall detection proposed in this paper is mainly realized through OpenMv’s built-in function library and the three-frame difference algorithm, and then the fluctuation range of the body center coordinates of the elderly is adjusted to determine whether the elderly has fallen.