基于心音信号分析的嵌入式身份识别系统开发

发布时间：2019-01-02 20:22

【摘要】：现代社会，随着交通、通讯和网络技术的飞速发展，人类的活动范围越来越大，身份识别的难度和重要性也越来越突出，对人类自身身份识别的准确性、安全性与实用性提出了更高的要求。所谓生物识别技术是指通过计算机与光学、声学、生物传感器和生物统计学原理等高科技手段密切结合，利用人体固有的生理特性和行为特征进行个人身份鉴定的技术。传统意义上的生物识别技术如指纹、虹膜、掌纹、语音等在实际应用中虽然取得了成功，但都面临着安全问题——使用者的身份易被仿冒，因此需要新的技术来克服这种不安全性和不可靠性。心音是来自于人体内部的内源性生理信号，所具有的独特生理特征使其有望成为新型身份识别方法之一。本文研究开发基于心音信号分析的嵌入式身份识别系统。在心音信号预处理、特征提取及匹配识别等关键技术领域研究基础上，重点设计了心音身份识别嵌入式系统的软硬件平台，开发心音身份识别软件，实现嵌入式心音身份识别系统。主要研究内容如下：（1）研究心音信号用于身份识别的原理及其时频域特性，回顾了心音身份识别的历史和国内外的研究现状。（2）提出了心音信号的小波包变换倒谱系数（Wavelet Package Cepstrum Coefficients，WPCC）特征参数提取方法，结合矢量量化生成稳定的表征个人身份的心音码本，并在LabVIEW平台上编程实现。（3）设计嵌入式心音身份识别系统的上下位机，下位机包括心音拾音头、有源滤波放大电路、主放大电路、dsPIC主控单元电路、HC-06蓝牙发送电路以及电源电路，上位机包括HC-06蓝牙接收电平转换电路、工控主板和液晶屏。下位机采集心音信号，对其进行滤波放大，经过AD转换后实现蓝牙无线传输；上位机接收心音信号，负责人机交互和识别算法处理，实现心音身份的有效识别。（4）基于LabVIEW设计心音身份识别软件，，结合MicrosoftAccess2007构建心音数据库，实现信号采集、用户注册、用户辨识和系统设置的功能，生成安装文件，实现在嵌入式平台上的无缝移植。（5）开发Windows Embedded Standard嵌入式操作系统，实现系统功能定制、驱动定制和启动定制，搭载基于LabVIEW设计的心音身份识别软件，灵活方便的实现基于心音信号的身份识别。本文的研究提供了一种准确率高、防伪能力强的生物识别新设备，可以为当前社会所面临的各种身份鉴定和信息安全问题提供较为理想的解决方案。
[Abstract]:In modern society, with the rapid development of traffic, communication and network technology, the range of human activities is becoming larger and larger, and the difficulty and importance of identity recognition are becoming more and more prominent. Safety and practicability put forward higher requirements. The so-called biometrics technology is a technology that combines computer with optics, acoustics, biosensor and biostatistics, and makes use of the inherent physiological and behavioral characteristics of human body to carry out personal identification. Traditional biometrics, such as fingerprint, iris, palmprint, speech and so on, have been successful in practical applications, but they all face security problems-the identity of the user is easy to be counterfeited. New technologies are therefore needed to overcome this insecurity and unreliability. Heart sound is an endogenous physiological signal from the human body, which is expected to be one of the new identification methods due to its unique physiological characteristics. In this paper, an embedded identity recognition system based on heart sound signal analysis is developed. Based on the research of heart sound signal preprocessing, feature extraction and matching recognition, the hardware and software platform of heart sound identity recognition embedded system is designed, and the heart sound identity recognition software is developed. The embedded heart sound identity recognition system is implemented. The main research contents are as follows: (1) the principle and time-frequency characteristic of heart sound signal used in identity recognition are studied, and the history of heart sound identification and the research status at home and abroad are reviewed. (2) the method of extracting feature parameters of heart sound signal based on wavelet packet transform cepstrum coefficient (Wavelet Package Cepstrum Coefficients,WPCC) is proposed. A stable heart sound codebook is generated by vector quantization and implemented on LabVIEW platform. (3) the upper and lower computer of the embedded heart sound identification system is designed. The lower computer includes the heart sound pickup head, the active filter amplifier circuit, the main amplifier circuit, the dsPIC main control unit circuit, the HC-06 Bluetooth sending circuit and the power supply circuit. Upper computer includes HC-06 Bluetooth receiving level conversion circuit, industrial control motherboard and LCD screen. The lower computer collects the heart sound signal, filters and amplifies it, realizes the Bluetooth wireless transmission after the AD conversion; the upper computer receives the heart sound signal, is responsible for the man-machine interaction and the recognition algorithm processing, realizes the heart sound identity effective recognition. (4) Design the heart sound identification software based on LabVIEW, build heart sound database with MicrosoftAccess2007, realize the functions of signal collection, user registration, user identification and system setup, generate installation files, and realize seamless transplantation on embedded platform. (5) develop Windows Embedded Standard embedded operating system, realize system function customization, drive customization and start customization, carry the heart sound identification software based on LabVIEW design, and realize the identity recognition based on heart sound signal flexibly and conveniently. The research in this paper provides a new biometric equipment with high accuracy and strong anti-counterfeiting ability, which can provide a more ideal solution to the problems of identification and information security in the current society.
【学位授予单位】：杭州电子科技大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：TP391.4;TP368.1

【参考文献】