基于PCA人脸识别的移动视觉应用

发布时间：2017-12-31 20:13

本文关键词：基于PCA人脸识别的移动视觉应用　出处：《吉林大学》2016年硕士论文　论文类型：学位论文

【摘要】：随着硬件技术和网络信息的发展,各种经济活动交流交往都涉及到安全问题,如何能够快速而又高效的进行权限认证是一个非常重要的问题。因此,一种用于人们身份识别的应用也就应运而生了,特别是在移动互联网时代,移动视觉应用的需求更大。在人体的多种生物特征中,人脸是人体具备的一种比较典型的生物特征,因此通过人脸进行身份识别,比其它方法更加方便、快捷、可靠。人脸识别是一个图像处理的过程,它选择包含人脸的图片来作为输入,并将该图片通过一定方式的处理,转化为向量,通过比较向量之间的相似程度来确定两幅图像的相似程度,并进一步根据阈值判定两幅图像是否是同一个类别,据此来判定目标人脸图像的身份。由于当前android平台的移动设备市场占有率高,价格便宜,在大众手里普遍取代专业相机来拍摄照片以及获取图像,本文选择安卓手机作为图像的获取的途径,并通过网络传输给电脑服务器端进行人脸识别,然后将结果通过网络反馈给手机应用端,以此实现普通大众都可以使用的的人脸识别。目前人脸识别的技术已经应用到各个行业之中,特别是安全防卫、金融、医疗设备以及在线教育等重要领域,其技术研究和实际应用都已经很成熟了。随着智能手机的快速发展,移动互联网的时代正式到来,互联网生活也从PC端转移到了移动端,各种应用程序如雨后春笋般冒出来,新闻阅读,在线交流,手机游戏,视频会话等功能越来越能够满足人们的日常生活。相信在不久的将来,随着硬件设备的进一步提高,很多工作也可以转移到移动端上面来。结合android手机的人脸识别的移动视觉应用就是将一些需要身边识别和权限认证的工作转移到移动端。这样的转变,既符合互联网生活的新潮流,也符合互联网+的新战略。基于此种考虑,本文选择结合移动设备的便捷和电脑端的性能做一个关于人脸识别的移动视觉应用,跟上互联网的脚步,也抓住互联网的机遇,做从PC端到移动端转变潮流中的一份子,并以此为契机探索和发展人脸识别在移动互联网时代的效果和作用。本文主要研究人脸识别中的特征提取和匹配识别两部分,并研究了与其相关的一些算法和方法。并根据研究的内容完成人脸识别系统的设计与实现,自建人脸图像库,完成人脸识别的过程。主要使用的方法是主成分分析法(Principle Component Analysis,PCA),对其进行介绍并实现。
[Abstract]:With the development of hardware technology and network information, the communication of various economic activities involves security issues. How to quickly and efficiently authenticate the authority is a very important issue. An application for human identity has emerged, especially in the era of mobile Internet, mobile vision applications are in greater demand. In the human body of a variety of biological characteristics. Face is a typical biological feature of human body. Therefore, face recognition is more convenient, faster and more reliable than other methods. Face recognition is a process of image processing. It selects the picture which contains the face as the input, and converts the image into vector by processing it in a certain way, and determines the similarity degree of the two images by comparing the similarity between the vectors. And then determine whether the two images are the same category according to the threshold value, and then determine the identity of the target face image. Because of the high market share of mobile devices in the current android platform, the price is cheap. In the hands of the general public instead of professional cameras to take photos and obtain images, this paper chooses Android phone as the way to obtain the image, and through the network transmission to the computer server for face recognition. Then feedback the results to the mobile phone application end through the network, so as to achieve the general public can use face recognition. At present, face recognition technology has been applied to various industries, especially security and defense, finance. Medical equipment and online education and other important fields, its technology research and practical applications have been very mature. With the rapid development of smart phones, the era of mobile Internet has come. Internet life has also shifted from PC to mobile, with applications springing up, news reading, online communication, mobile games. Video session and other functions can more and more meet people's daily life. I believe that in the near future, with the further improvement of hardware devices. A lot of work can also be transferred to the mobile end. The mobile vision application of face recognition combined with android mobile phone is to transfer some work that requires side recognition and authority authentication to the mobile side. Change. It is not only in line with the new trend of Internet life, but also in line with the new strategy of the Internet. Based on this consideration, this paper chooses to combine the convenience of mobile devices and the performance of computer to make a mobile vision application about face recognition. Keep up with the pace of the Internet, but also seize the opportunities of the Internet, from the PC to the mobile end of the shift in the trend of a part. And take this opportunity to explore and develop the effect and role of face recognition in the mobile Internet era. This paper mainly studies the feature extraction and matching recognition in face recognition. And some related algorithms and methods are studied, and according to the content of the study, the design and implementation of the face recognition system is completed, and the face image database is built. The main method is principle Component Analysis (PCA), which is introduced and implemented.
【学位授予单位】：吉林大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TP391.41

【相似文献】