人体姿态检测系统的开发及智能服装的设计  

Development of Human Posture Detection Systemand Design of Intelligent Clothing

作　　者：王世豪 祝双武 臧衍乐 马晓彤 马阿辉 李丑旦 WANG Shihao;ZHU Shuangwu;ZANG Yanle;MA Xiaotong;MA Ahui;LI Choudan(College of Textile Science and Engineering,Xi’an Polytechnic University,Xi’an Shaanxi 710048,China)

机构地区：[1]西安工程大学纺织科学与工程学院,陕西西安710048

出　　处：《传感技术学报》2024年第8期1342-1349,共8页Chinese Journal of Sensors and Actuators

摘　　要：针对人体久坐易影响心脏等器官、跌倒救助不及时容易危及生命等姿态异常带来的健康安全问题,采用单片机STM32F103C8T6和六轴运动处理传感器MPU6050等模块设计了一款能够进行姿态检测的智能服装,通过卡尔曼滤波对MPU6050采集的数据进行处理,采用合加速度、角度的阈值处理方法判断用户的身体姿态,为了提高姿态检测的准确性,同时采集用户手机内置的加速度、角速度传感器检测的数据进行辅助判断。实验结果表明,结合手机传感器的数据后,该智能服装对久坐和跑步的姿态识别准确率达到了99.3%和96.7%;对跌倒和下蹲的识别准确率为98%和95.3%,对姿态的识别具有很高的准确性,且健康提醒和安全报警及时,能够有效保障穿戴者的身体健康。Abnormal posture can bring health risks to our body,e.g.,sedentariness easily affects multiple organs such as heart,and untime-ly fall assistance might even be life-threatening.To cope with those issues,an intelligent garment capable of performing posture detection is developed with single chip of STM32F103C8T6 and six-axis motion processing sensor of MPU6050.The data collected by MPU6050 are processed through Kalman filtering and then used for posture detection by employing the threshold processing method of acceleration and angle.Additionally,the data collected by the built-in acceleration and angular velocity sensors of the carry-on mobile phone are called to improve the accuracy of body posture detection.The experimental results show that after combining the data of the mobile phone sensors,the developed intelligent garment achieves 99.3%and 96.7%recognition accuracy for long periods of sitting and running,respectively,and regarding falls and squats,the achieved recognition accuracies are 98%and 95.3%,respectively.With high posture detection accuracy and timely health reminders and safety alarms,the developed intelligent garment can effectively protect the health of the wearers.

关 键 词：智能服装 姿态检测 报警定位 卡尔曼滤波 安卓手机APP 

分 类 号：TS941.26[轻工技术与工程—服装设计与工程]