人脸识别系统的实现解析.docVIP

目录 1 绪论 1 1.1 课题背景 1 1.1.1 课题来源 1 1.1.2 人脸识别技术 1 1.1.3 人脸识别技术的研究意义 2 1.2 人脸识别发展历史与概况 3 1.3 人脸识别的难点 4 2 人脸识别的相关理论 1 2.1 人脸特征 1 2.1.1 肤色特征 1 2.1.2 灰度特征 1 2.2 PCA的基本概念 1 2.2.1 PCA算法的原理 2 2.2.2 PCA算法 3 3 人脸识别算法设计 1 3.1 人脸检测与定位 1 3.2 基于PCA的人脸特征提取 2 3.3 人脸识别方法 3 3.3.1 几何特征法 3 3.3.2 特征脸法 4 3.3.3 弹性图匹配法 5 3.3.4 神经网络法 6 4 基于PCA的人脸识别系统设计 1 4.1 引言 1 4.2 具体算法 1 4.3 ORL人脸库实验分析 2 4.3.1实验1：算法的比较 3 4.3.2 实验2：样本数目的选择 4 4.3.3 实验3：采样因子的确定 4 4.4 YALE人脸库实验 5 5 小结 1 5.1 总结 1 5.2 展望 1 致谢语 1 参考文献 1 附件 1 摘要：人脸识别技术是一种新型的身份识别技术，它是生物识别技术的一种。随着人们应用需求的不断深入，人脸识别技术将呈现出主流化、芯片化、标准化发展趋势。人脸识别系统主要由检测与定位、特征提取与识别两部分组成，在这两部分的基础上还包括预处理等步骤对于一个全自动的人脸识别系统，整个的识别过程可归纳为人脸检测与定位、特征提取和识别实现流程如下：首先是预处理，对图像进行光照处理等以改善图像质量；检测与定位，从不同场景中检测出人脸并将其从背景中分割出来；对人脸库中所有图像大小和各器官位置归一化；最后对归一化的人脸图像进行特征提取与识别。由于动态识别研究处于起步阶段该技术相对较欠缺，因此本文只研究基于静态人脸图像的识别方法。关键词：人脸识别；检测方法Abstract: Face recognition technology is a new identity, it is a kind of biometric technology, face recognition technology is also a very active research area, which covers digital image processing and many other fields. With the deepening of peoples application requirements, the face recognition technology will appear in the mainstream, chip, standardization, development trends. Face recognition system mainly by the detection and location, feature extraction and recognition of two parts, including the pretreatment steps in these two parts on the basis of a fully automatic face recognition system, the whole recognition process can be summarized as the human face detection and location, feature extraction and recognition. The process is as follows: First, pretreatment, light processing to improve image quality; detection and location on the image to detect a persons face from different scenes and split out from the background; face database image size and each organ location normalization; Finally, the normalized face image feature extraction and recognition. Dynamic discrimination in its infancy, the technology is relatively lacking This article only study identification me