基于嵌入式系统人脸识别方法的研究

发布时间：2018-07-16 08:08

【摘要】：随着社会的不断进步以及各方面对于快速有效的自动身份验证的迫切要求，生物特征识别技术在近几十年中得到了飞速的发展。人脸识别技术作为生物识别技术之一，是模式识别领域中一个热点研究问题。人脸识别是指基于现有的人脸样本集，利用一定的算法，提取人脸图像中的人脸视觉特征信息，运用图像处理和模式识别等技术从图像或者视频中获得并分析一个或多个人脸，然后从中提取出有效的识别信息，来自动鉴别图像中待识别人的身份的一门技术，它涉及计算机图形学，计算机视觉、模式识别、机器学习、感知科学、人工智能、计算机智能等多门学科技术。人脸识别与其他传统生物识别技术相比,具有便于采集、方便快捷、交互友好等优点,已逐渐被大众接受，在智能人机接口、图像检索、视频处理、安全等领域都有着极为广泛的应用价值。从广义上说，人脸识别主要有两个部分：人脸检测与人脸识别。在构建人脸识别系统中，人脸识别技术包括人脸图像采集，人脸以及人眼定位，人脸图像识别预处理，特征提取，，身份识别确定等几个方面。本文所要研究的人脸识别是在狭义范围内研究人眼定位后，人脸图像预处理，特征提取后，进行身份识别判断。本文所做的工作主要有介绍了人脸识别国内外研究背景以及研究意义，人脸识别理论及算法以及图像处理算法进行了介绍。接下来对嵌入式系统进行设计，选用三星公司的S3C2440ARM微处理器作为硬件基础，并进行了Linux内核移植，U-boot移植等嵌入式操作系统的搭建工作，最后采用主成分分析（PCA）算法对人脸特征进行提取，运用Adaboost算法对样本集进行训练，用欧式距离衡量特征匹配检验了研究结果。
[Abstract]:With the development of society and the urgent requirement of rapid and effective automatic authentication, biometric identification technology has been developed rapidly in recent decades. As one of the biometric techniques, face recognition is a hot topic in the field of pattern recognition. Face recognition is based on the existing face sample set, using certain algorithm to extract face visual feature information from face image, using image processing and pattern recognition technology to obtain and analyze one or more human faces from image or video. And then extract effective recognition information from it to automatically identify the identity of the person to be identified in the image, which involves computer graphics, computer vision, pattern recognition, machine learning, perceptual science, artificial intelligence, Computer intelligence and other multi-disciplinary technology. Compared with other traditional biometrics, face recognition has the advantages of easy collection, convenience and friendly interaction, and has been gradually accepted by the public, such as intelligent man-machine interface, image retrieval, video processing, etc. Safety and other fields have extremely wide application value. Broadly speaking, face recognition has two main parts: face detection and face recognition. In the construction of face recognition system, face recognition technology includes face image acquisition, face and human eye location, face image recognition preprocessing, feature extraction, identity identification and so on. In this paper, face recognition is studied in a narrow sense after the study of human eye location, face image preprocessing, feature extraction, identity recognition judgment. This paper mainly introduces the research background and significance of face recognition at home and abroad, the theory and algorithm of face recognition and image processing algorithm. Then the embedded system is designed, the Samsung S3C2440arm microprocessor is chosen as the hardware base, and the embedded operating system, such as transplanting Linux kernel to U-boot, is built. Finally, principal component analysis (PCA) algorithm is used to extract face features, Adaboost algorithm is used to train the sample set, and Euclidean distance measurement feature matching is used to test the research results.
【学位授予单位】：青岛科技大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：TP391.41;TP368.1

【参考文献】