人体和非人体PIR信号识别.doc

人体与非人体PIR信号的AR识别模型研究 重庆大学士学位论文 学生姓名：贺莉芳 指导教师：龚卫国 教 授 专　　业：仪器科学与技术 学科门类：工 学 重庆大学工程学院O一一年四月  Research on AR Recognition Model of Human and Non-human PIR Signal A Thesis Submitted to Chongqing University in Partial Fulfillment of the Requirement for the Degree of Master of Engineering By He Lifang Supervised by Prof. Gong Weiguo Major：Instrument Science and Technology College of Opto-electronic Engineering of Chongqing University, Chongqing, China April, 2011  摘 要、稳定。PIR信号数据库的建立。为采集数据首先研制一套PIR信号采集装置，采用通用的PIR探测器（RE200B型PIR传感器）对探测区域的人体和非人体红外热源进行数据采集。分别采集人体随机运动路径、随机速度数据；非人体（不同大小的狗、鸡等）随机运动路径、随机速度数据；在不同环境不同天气下的人体和非人体数据。根据采集的数据建立人体PIR信号与非人体PIR信号的数据库，作为数据分析的基础。 2）PIR信号处理方法的研究。根据PIR信号特征，研究了基于AR模型特征提取的方法，并根据提取的特征对PIR信号进行人体和非人体识别。该算法先对原始信号进行预处理去噪，根据AR模型阶数选择原则确定模型阶数，利用BURG算法估算AR模型系数作为信号特征。设计支持向量机（SVM）（NN） ABSTRACT Nowadays in some countries and regions, there have been increasing investments in the security system in public place and residential area with the rapid development of security market. Pyroelectric infrared (PIR) detectors can detect the mobile infrared radiation sources in the sensitive areas and implement human movement detection at normal temperature. It is so widely used in security fields because of its low cost performance, conveniences to be concealed, stable performance and no illumination requirement. It is one of the most common monitoring products in invasive alarm systems. However, there are serious false alarms and missed alarm phenomenon in the universal PIR detectors at present, so, PIR detector faced a serious challenge in the popularization of security construction. The main reason of false alarm and missed alarm is the lack of using the signal processing technology to analyze the signals with in-depth researching PIR detector principle. Therefore, improving the PIR detector performance base on in-deep analysis of signals with signal processing technology and designing a low false and missed alarm rate intelligent PIR detector is very impor