基于三轴加速度计和SVM算法的校园运动识别  

Campus movement recognition based on three-axis accelerometer and SVM algorithm

作　　者：周露 宋浩兰 白静蕾 闻家豪 李继亮 ZHOU Lu;SONG Haolan;BAI Jinglei;WEN Jiahao;LI Jiliang(School of Electronic Engineering,Xi’an Shiyou University,Xi’an 710065,China)

机构地区：[1]西安石油大学电子工程学院,陕西西安710065

出　　处：《电子设计工程》2022年第21期80-84,共5页Electronic Design Engineering

摘　　要：人体运动行为识别在医疗、安全、健康等方面有着广泛应用,为了探究如何高效准确获取人体运动行为信息,提出一种基于三轴加速度传感器与支持向量机的大学生运动行为识别算法。该方法通过佩戴在皮带扣处的三轴加速度传感器,采集大学生运动行为,运动行为包括走路、站立、跑步、躺卧、上楼、下楼六种状态。通过支持向量机实现识别模型的建立与预测,得到总体分类准确度为93.84%。实验结果表明,基于三轴加速度传感器信号与支持向量机算法能够较好地识别大学生在校园的运动行为。Human motion behavior recognition is widely used in medical treatment,safety,health,etc.In order to explore the efficient and accurate acquisition of human motion behavior information,a college student motion behavior recognition algorithm based on three-axis acceleration sensor and support vector machine is proposed. In this method,a three-axis acceleration sensor worn on the belt buckle is used to collect the sports behaviors of college students. The sports behaviors include six states of walking,standing,running,lying,going upstairs,and going downstairs. Through feature extraction,a motion data set is formed,and a support vector machine recognition model is obtained through training. An overall classification accuracy of 93.84% is achieved. The experimental results show that based on the signal of the three-axis acceleration sensor and the support vector machine algorithm,the sports behavior of college students on campus can be better recognized.

关 键 词：三轴加速度传感器 模式识别 特征提取 支持向量机 

分 类 号：TN609[电子电信—电路与系统]