首页 | 官方网站   微博 | 高级检索  
     

基于PIR Sensor的单目标跟踪系统的设计与实现
引用本文:李博雅,李方敏,刘新华,熊迹,李等.基于PIR Sensor的单目标跟踪系统的设计与实现[J].传感技术学报,2014,27(9).
作者姓名:李博雅  李方敏  刘新华  熊迹  李等
作者单位:武汉理工大学信息工程学院,武汉,430070
摘    要:针对基于热释电红外传感器的人体跟踪系统在实际环境中,由于受环境噪声和硬件参数影响而误差较高的问题,本文提出了一种基于热释电红外传感器PIR Sensor(Pyroelectric Infrared Sensor)的高精度人体目标跟踪方案。该方案首先对检测区域内运动人体热辐射的红外信号特征进行了提取,然后利用PIR Sensor定位节点自身几何参数和探测数据,得到初步定位结果。最后通过Kalman滤波算法对初步定位结果进行滤波处理并更新目标的状态信息,实现对检测区域内人体目标的定位与跟踪。实验表明,该系统的跟踪误差与同类跟踪系统相比降低了71.96%,证明了该系统具有较高的跟踪精度。

关 键 词:目标跟踪  信号提取  热释电红外传感器(PIR  Sensor)  Kalman

Design and Implementation of Single Target Tracking System Based on PIR Sensor
LI Boya,LI Fangmin,LIU Xinhua,XIONG Ji,LI Deng.Design and Implementation of Single Target Tracking System Based on PIR Sensor[J].Journal of Transduction Technology,2014,27(9).
Authors:LI Boya  LI Fangmin  LIU Xinhua  XIONG Ji  LI Deng
Abstract:Due to the effects of the environmental noise and hardware parameters, the tracking error rate based on the PIR sensor is relatively higher in the real environment. For this point, this paper proposed a high accuracy human tracking scheme that based on pyroelectric infrared (PIR) Sensor. Firstly, PIR sensors detect the infrared signal of the moving human in the sensing area and the characteristics of the infrared signal are extracted. Then, the geometric parameters of locating node and the detecting data are used for the preliminary positioning result. Finally, Kalman filter algorithm is used to filter the preliminary positioning result and update the target state information for realizing the human target positioning and tracking in the sensing area. Experiments show that the tracking error has decreased by 71.96% comparing with other similar tracking systems. It has been proved that the system has a high tracking accuracy.
Keywords:target tracking  signal extracting  pyroelectric infrared( PIR) Sensor  Kalman EEACC:7230  7320R
本文献已被 CNKI 万方数据 等数据库收录!
点击此处可从《传感技术学报》浏览原始摘要信息
点击此处可从《传感技术学报》下载全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司    京ICP备09084417号-23

京公网安备 11010802026262号