一种基于移动终端的新型计步方法

A New Step Detection Approach Based on Mobile Termination

近年来,随着微电子技术和计算技术的发展以及智能手机和穿戴设备的普及,生物信号处理以及模式识别成为工程领域的热门话题.由于中国人口老龄化,适宜的身体锻炼和健康医疗已经成为社会关注的热点.计步器作为一种运动检测设备进入到人们生活中,同时智能手机上有计步功能的应用软件得到普及,但是目前的计步算法不能很好地去除人们生活中产生的噪声,影响计步精度,该文提出了一种高精度计步方法,目标是去除计步算法中的噪声,减少其他因素对计步的影响.该计步方法基于智能手机中加速度传感器的三维离散信号,对三维信号进行分析,提取信号中的特征,最终高精度地统计人行走的步数.该文首先对加速度传感器三维信号的选取进行了讨论,采取平滑滤波算法对信号进行去噪,接着提取信号中的特征并使用M5算法对信号进行分类,最终对有效信号采取动态时间归整(Dynamic Time Warping,DTW)算法进行计步.该文最后对此计步方法的精度和抗干扰能力进行评测,证明该方法在统计步数上具有较高的精度和抗干扰能力.

In recent years,with the development of microelectronics and computing technology,the popularization of smartphone and wearable devices,biological signal processing and pattern recognition became a hot topic in the field of engineering.Due to the aging population in China,physical exercise and proper health care have attracted a hot social attention.The pedometer as a movement detecting device has entered into people＇s lives.At the same time,the smartphone equipped with a pedometer becomes more and more popular.But the current step algorithm cannot well remove the noise from people＇s lives which impacts the precision of pedometer.This paper proposes an approach of high-precision pedometer,whose main goal is to eliminate the effect of noise and reduce the influence of other factors.The approach is based on the threedimensional discrete signal of accelerometer in the smartphone.After processing and analyzing the three-dimensional signal,extract features of the signal,the number of steps can be obtained more accurately at the end.In this paper,firstly,the selection of three-dimensional acceleration sensor is discussed,and the smoothing algorithms are adopted to eliminate signal noise.Secondly,the features extracted from the signals are classified by M5 algorithm.Thirdly,for effective signals,dynamic time warping algorithms are used to count steps.Finally,we evaluate and verify the higher precision and anti-jamming capability of this approach.