【作者】 王宪保；

【导师】 王守觉；

【作者基本信息】 浙江工业大学 ， 控制理论与控制工程， 2003， 硕士

【摘要】 人脸自动识别技术有着广阔的应用前景和迫切的现实需求，是当前模式识别领域最热门的研究方向之一。本文首先对人脸识别技术的研究内容和人脸识别技术的发展做了概括性的介绍，接着对人脸检测和识别技术的研究现状进行了较全面的论述，然后介绍了神经网络技术在人脸识别中的应用，最后详细叙述了基于仿生模式识别的人脸身份确认系统。 本文从以下几个方面对人脸识别技术做了探讨： 1．通过定义图像块的复杂度，快速而又准确地找到人脸图像中的眼睛、鼻子和嘴巴等特征器官，为人脸的姿势校正，特征参数的提取提供了有力的保证。 2．通过寻找各个特征器官的中心，进而提取特征器官周围的细微信息的方法来提取人脸的特征参数，提出了一种新的人脸特征参数提取方法。 3．用三个摄像头摄取图像弥补了用单像头摄取二维图像表达三维人脸的不足，使得通过融合同一时刻采取的多幅图像更加完备地表达同一人脸。 4．运用仿生模式识别理论建立起识别神经网络。仿生模式识别从“认识”而不是“区分”的角度来进行模式识别，是对学习样本的全体在特征空间中形成的复杂几何体“形状”的分析和“认识”，为用高维空间理论建立识别神经网络提供了理论基础。更多还原

【Abstract】 Automatic human face recognition is attractive in pattern recognition and image processing and has a broad prospect in the application in reality.In this paper, first, the content and development of automatic human face recognition are introduced briefly; then, a survey of automatic human face detection and recognition is given; second, the application of neural network in automatic human face recognition is talked about; last, it is discussed in detail about the system of human face personal identification based on the biomimetic pattern recognition.In this paper several views are given as follows:1. In this paper, the feature are found rapidly and rightly through the definition of the complexity of picture . It provides a strong support for the adjustment of posture and the feature extraction.2. A new method of the feature extraction from human face is presented. It firstly finds out the centers of organ and the subtle of the features are considered as face features.3. In this paper, three video cameras are used in order to describe faces more completely than two video cameras.4. The neural network based on the biomimetic pattern recognition principles is built. The biomimetic pattern recognition makes recognition from the views of "matter cognition" instead of "matter classification", which analyzes and cognizes the high dimensional geometrical distribution that consists of the sample sets in the high dimensional feature space. It provides the theoretic basis of building the neural network based on high dimensional theory.更多还原