摘要步行和慢跑对人体健康的益处毋庸置疑，随着人们运动观念的增强以及对步行认识的进步，近两年，掀起了一股健身的热潮，计步的准确性也因此受到人们的关注。纵观现在市面上的计步产品，大多采用手机或者佩戴式设备内的三轴加速度传感器计步，由于人体活动复杂，容易发生误计的情况，造成计步误差。86166

本文以准确计步为主要目标，提出了基于压力传感器、六轴传感器的多参数计步算法，综合利用了嵌入式系统技术、传感器技术、蓝牙技术、Android技术，开发出一套智能健康计步器系统。系统分为数据采集端和手机监控终端。数据采集端采集压力传感器和六轴传感器的运动参数，通过对运动参数的处理分析，经SVM算法判别脚上抬和脚下落两个动作，获得运动步数后，通过蓝牙发送给手机监控终端，由监控终端将运动数据反馈给用户，并结合运动提醒等辅助功能，帮助用户智能健康地运动。

本文先介绍了系统所用到的相关技术，再结合系统功能定义给出了系统总体方案的设计，并详细阐述了各个模块的软硬件设计，最后通过压力传感器单独计步、多传感器多参数融合计步的计步准确率、误判率对比分析，验证了基于压力传感器、六轴传感器的多参数计步具有更高的准确度。

毕业论文关键词：计步；人体行为识别；分类识别算法；传感器；嵌入式系统

Abstract There is no doubt that walking ang jogging is good for human health。 With the enhancement of people's sense of movement and the progress of  knowing the importance of walking,  the fitness upsurge is setting off in recent years, and the accuracy of step catch people's attention following。 Looking at the current market of step counter products, most of the mobile or wearable devices counter steps in three-axis acceleration sensor。 Due to the complexity of the human’s activities, error count will happen easily, which resulting in the step counting error。

Counting step accurately is the main target in this this paper, whith proposing a step counter algorithm, which is based on pressure sensor and multi parameters of six axis sensor。 With the comprehensive utilization of the embedded system technology, sensor technology, Bluetooth technology, Android technology, developed a set of intelligent health pedometer system。 The system is pided into a data acquisition terminal and a mobile monitoring terminal。 The data acquisition terminal is response for collecting the pressure sensor’s and six sensor’s data。 Through the processing and analysis of the motion parameters , discriminate foot lift and foot falls actions according to the SVM algorithm。 Then the data acquisition terminal send the steps to the mobile phone terminal with Bluetooth。 The monitoring terminal will motion data feedback to the user, and combining the movement to remind, recommendations and other auxiliary functions, helping users exercising more intelligency and healthy。

Firstly, this paper introduces the related technologies used in the system, combined with the system function definition giving the design of the overall scheme of the system, and explains the design of each modulein of the hardware and software detaily。 Through comparing the step count quasi accuracy and error rate of the pressure sensor single step and multi sensor multi parameter fusion meter step, verificate that the multi parameter step counting has the higher step counting accuracy。

Keywords: step counter; human behavior recognition; classification and recognition algorithm; sensor; embedded system

目  录

第一章 绪论 1

1。1 研究背景及意义 1

1。2 国内外发展现状