基于Web网络的身份识别系统的设计与实现

发布时间：2018-03-22 03:16

本文选题：web网络　切入点：身份识别　出处：《微电子学与计算机》2017年05期 　论文类型：期刊论文

【摘要】：当前身份识别系统大多采用密码识别技术,容易被窃取或遗失,导致用户有一定的损失,识别精度较低.为此,设计了一种新的基于web网络的身份识别系统,主要由浏览器端录音模块、声纹数据传输模块和声纹训练及识别模块构成.给出系统架构,依据系统架构在浏览器端安装一个Activex录音控件,对用户的语音进行录制.通过预加重、端点检测等方式对声纹信号进行预处理.对声纹信号进行特征值提取,通过信号转换过滤信号中没有用的部分,提取出能够体现用户录制语音本质属性的特征参量,依据提取特征值得到D值,通过D值实现身份识别.实验结果表明,所设计系统有很高的识别精度.
[Abstract]:Most of the current identity recognition systems use cryptographic identification technology, which is easy to be stolen or lost, resulting in a certain loss of users and low recognition accuracy. Therefore, a new identification system based on web network is designed. It is mainly composed of browser end recording module, voiceprint data transmission module and voice pattern training and recognition module. The system structure is given. According to the system architecture, a Activex recording control is installed on the browser end. Record the voice of the user. Preprocess the voiceprint signal by pre-weighting and endpoint detection. Extract the eigenvalue of the voiceprint signal and filter the useless part of the signal by signal conversion. The characteristic parameters which can reflect the essential attribute of the speech recorded by the user are extracted and the D value is obtained according to the extracted characteristic value. The experimental results show that the system designed has a high recognition accuracy.
【作者单位】：中国劳动关系学院计算机应用技术教研室;
【分类号】：TP393.092

【相似文献】