Opencv+Adaboost基于人脸识别与认证的准入系统设计
时间:2018-03-02 14:41 来源:毕业论文 作者:毕业论文 点击:次
| 摘要近些年来，人脸识别是计算机视觉和模式识别领域的研究热点之一，陆续出现了一些人脸识别商业系统应用于公众和个人安防等方面。广义地讲，人脸识别系统一般包括人脸检测，预处理，特征选择与提取，人脸匹配四个步骤。本文以人脸识别的实用化，实时化和系统化为目标，对实用人脸识别系统中若干问题进行了初步探讨。论文的主要工作如下： 1. 使用基于Haar特征的Adaboost人脸检测算法检测测试图像中的人脸。由于AdaBoost算法主要利用人脸的灰度特征，而摄像头捕捉的图像为彩色图像，因此在检测之前需要先将彩色图像转化为灰度图像。该算法特征提取速度快，而且计算简单。10928 2. 对检测出的人脸图像进行预处理操作，主要为几何归一化和光照预处理。 3. 对进行预处理操作后的标准人脸图像进行特征抽取，本文先后实现了PCA算法和LBP算法，并对两种算法进行了比较。 4. 实现基于人脸识别与认证的准入系统，本文以MFC和Opencv为技术平台，开发环境为VS2010和Opencv2.3.1。 关键词 人脸识别 人脸检测 Adaboost算法 预处理 PCA LBP 毕业设计说明书（论文）外文摘要 Title Access System Based On Face Recognition And Authentication Abstract Recent years, face recognition has become a hot topic of computer vision and pattern recognition. And some commercial face recognition systems have been developed and applied in public and inpidual security. In general, a face recognition system is accomplished in four steps, i.e. face detection, preprocessing, facial feature selection and extraction, face recognition. In this thesis, some key issues are primarily studied, aiming at building real-time practical face recognition system. And the main work of this thesis can be described as follows: 1. Use Adaboost face detection algorithm based on Haar features to detect faces in test images. Because of that adaboost algorithm mainly use gray features of faces, while the camera captures the color image, we should convert color image into gray image first before detection. It is fast for feature extraction and simple for calculation. 2. Operate preprocessing onto the detected face image, including geometry normalization and illumination normalization. 3. Operate feature extraction onto the standard face image obtained after preprocessing. This thesis has realized the PCA (Principle Component Analysis) algorithm and LBP (Local Binary Pattern) algorithm respectively. And make comparisons of the two different algorithms. 4. Realize the access system based on face recognition and authentication .This thesis uses MFC(Microsoft Foundation Class) and Opencv(Open Computer Vision) as technical platforms, and uses VS2010 and Opencv2.3.1 as development environment. Keywords Face Recognition, Face Detection, Adaboost Algorithm, Preprocessing, PCA, LBP 目 次 1 绪 论 3 1.1 人脸识别的研究背景和意义 3 1.2 人脸识别的发展历史及国内外现状 3 1.2.1 发展历史 3 1.2.2 国内外现状 4 1.3 人脸识别的研究内容与研究方法 5 1.3.1 研究内容5 1.3.2 研究方法 6 1.4 论文的研究内容及章节安排 7 2 基于Adaboost算法的人脸检测 8 2.1 Boosting算法的基本原理 8 (责任编辑：qin) |