基于人脸识别的远程认证系统的研究与设计

发布时间：2018-08-11 21:30

【摘要】：人脸识别技术作为一种热门的生物特征识别技术,具有非常高的识别准确率和不易仿造性而被广泛应用。人脸识别技术已经被广泛的应用在各个与人们生活息息相关的领域,其中包括：社会保障福利、电子商务、安全防务、银行、等各个领域。随着移动互联网的快速发展,将移动互联网技术和人脸识别技术有机的结合在一起将是人脸识别技术应用和研发的一个新的方向。例如：现在阿里巴巴马云所提出的移动刷脸支付系统,就是利用人脸识别技术进行银行支付确认的应用,代替了传统的密码,具有方便、安全、快捷的优点。本文所提到的远程人脸认证系统,正是结合人脸识别的优势和智能手机的便携性,利用远程传输技术,对入脸进行远程识别认证,根据认证的结果而赋予一定的权限,通过远程认证的方法实现权限认证。本研究可以应用在多个场合,如：远程公司数据系统登录认证、远程认证人脸防冒领养老金、远程人脸追踪识别等场合都可以使用。远程人脸识别作为一种新型的研究技术,能非常好的适应当代社会各项发展当中,并且人脸识别技术拥有广阔发展前景和重大的意义。本课题主要对远程数据传输人脸识别方法进行深入讨论。主要研究内容如下：(1)研究分析人脸预处理方法,主要对本文所提出的人脸预处理方法的理论进行研究分析,对Haar-like特征和积分图方法进行人脸检测算法进行了深入分析,同时对人脸几何规范化方法、灰度规范化的过程和要求进行介绍,并给出了详细计算方法。(2)人脸特征识别技术的算法主要由两大类组成,分别是基于整体的人脸识别算法和基于局部的人脸识别算法。本文所采用的是基于整体特征脸的识别算法。深入研究了Fisherface算法,Fisherface算法是将PCA和Fisher准则相结合的一种改进算法,并分析了Fisherface算法在不同的环境和条件下的人脸识别效果。(3)实现远程数据传输的方法有很多种,本文采用了Http通信的方式,采用这种方式具有连接可靠、使用方便的特点。可以在安卓智能手机客户端和服务器之间进行图片信息的传输,本文详细研究和分析了实现通信的方式和保障通信可靠连接的方法。
[Abstract]:As a popular biometric recognition technology, face recognition is widely used because of its high recognition accuracy and difficulty in imitation. Face recognition technology has been widely used in various fields closely related to people's lives, including: social security benefits, e-commerce, security and defense, banking, and other fields. With the rapid development of mobile Internet, the combination of mobile Internet technology and face recognition technology will be a new direction in the application and development of face recognition technology. For example, the mobile brushing payment system proposed by Alibaba Ma Yun is an application of bank payment confirmation using face recognition technology, which replaces the traditional password and has the advantages of convenience, security and speed. The remote face authentication system mentioned in this paper is based on the advantages of face recognition and the portability of smart phone. The authority authentication is realized by the method of remote authentication. This study can be used in many situations, such as remote company data system login authentication, remote authentication of face anti-counterfeiting pension, remote face tracking recognition and so on. As a new research technology, remote face recognition can adapt to the development of contemporary society, and face recognition technology has broad prospects and great significance. This paper mainly discusses the method of face recognition based on remote data transmission. The main research contents are as follows: (1) face preprocessing method is studied and analyzed. The theory of face preprocessing method proposed in this paper is studied and the face detection algorithm based on Haar-like feature and integral graph is analyzed deeply. At the same time, the geometric normalization method of face, the process and requirements of gray level normalization are introduced, and the detailed calculation method is given. (2) the algorithm of face feature recognition mainly consists of two categories. Face recognition algorithm based on the whole and local face recognition algorithm respectively. The recognition algorithm based on global feature face is used in this paper. This paper deeply studies the Fisherface algorithm, which is an improved algorithm combining PCA and Fisher criterion, and analyzes the face recognition effect of Fisherface algorithm in different environments and conditions. (3) there are many methods to realize remote data transmission. In this paper, Http communication mode is adopted, which has the characteristics of reliable connection and convenient use. It can transmit picture information between the client and server of Android smart phone. This paper studies and analyzes the way of realizing communication and the method of ensuring reliable connection of communication in detail.
【学位授予单位】：内蒙古大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TP391.41

【参考文献】