卡尔曼滤波在人体姿态检测中的应用  

The application of Kalman filter in body posture detection

作　　者：杜波[1] 张卓[1] 刘垚[1] 

机构地区：[1]长春工程学院电气与信息工程学院,长春130012

出　　处：《长春工程学院学报（自然科学版）》2015年第3期98-103,共6页Journal of Changchun Institute of Technology：Natural Sciences Edition

基　　金：吉林省教育厅项目(120140059;12014K013)

摘　　要：为了尽早发现老人的意外跌倒,避免长躺现象,争取救护时间,对人体姿态进行检测。由控制器STM32F103和传感器MPU6050构成信号采集系统,采用无线蓝牙模式进行数据传递,由接收系统进行数据分析和显示。建立人体坐标系,对采集的加速度信号进行卡尔曼滤波,调节参数,补偿测量误差,实现最优估计。对滤波后的加速度信号进行整合,并以角度信号为辅助,进行姿态判断,及时判断跌倒事故。通过Matlab仿真发现,卡尔曼滤波的使用有效地提高了系统的测量精度,有助于姿态判断的实现。Human body posture should be detected in order to find senior people falling accident early and a‐void long time lying‐down ,also for saving time for emergency treatment .The signal collecting system composed of controller STM32F103 and sensor MPU6050 uses wireless bluetooth mode to make data transmission .The receiving system can make data analysis and display ,and set up the human body coordi‐nates .The system can compensate the measurement error and get optimal estimating for collected accelera‐tion signals that through Kalman filtering and parameters adjustment .It can analyze posture and detect falling accident earlier by integration of filtered acceleration signal , along with Angle signal assist . Through Matlab simulation test ,the application of Kalman filtering can improve system measurement pre‐cision effectively and be helpful for detection to body posture .

关 键 词：卡尔曼滤波 姿态检测 加速度信号 MATLAB 蓝牙 

分 类 号：TP274[自动化与计算机技术—检测技术与自动化装置]