基于图像处理的人眼识别方法的解析.doc

基于图像处理的人眼识别方法的研究 摘 要 人脸的识别是人类视觉最杰出的能力之一，人脸也具有和指纹一样的唯一性,也可以用来鉴别人的身份。眼睛与人脸其他部分相比有着灰度变化十分明显的特征，作为人脸部重要组成部分，眼睛的检测定位对于人脸信息的处理具有重要的作用与意义。 首先为了在一定程度上消除图像的噪声，并降低光照变化对人眼识别的影响，运用几种常用的图像的预处理方法来对所采集来的图像进行预处理，并且详细描述了它们的计算过程，展示具体的预处理效果。其次通过对几种常用于人脸的色彩空间的肤色识别算法的学习研究以及对比分析，从图像中分离出人脸。在实现正确人脸检测的基础上，利用积分投影算法实现人眼两个中心点的定位，完成人脸图像的归一化处理。为更准确的定位中心点，采取了适当的图像预处理技术，包括图像增强技术和图像的二值化处理技术，使得积分投影曲线在一定程度上免除噪声的干扰。 关键词：图像预处理，人脸检测，人眼识别，色彩空间，积分投影 Study of uman?Recognition?Method Based on Image Processing ABSTRACT Face recognition?is one of the?outstanding ability?of human vision. The face is a unique as fingerprints, can also be used to distinguish the identity of the person. Eyes?compared with other?part of the face?with the?gray scale change?is very obvious,?as an important part of?human face?detection and location?of eyes，plays an important role and significance?for the?processing of face?information. First of all, in order to eliminate the image noise to a certain extent, and reduce the influence of illumination change on human recognition, the pretreatment methods of image to the image pre-processing, and a detailed description of the calculation process of them, showing the effect of pretreatment of concrete. Secondly, through the color recognition of several commonly used in the face of the color space algorithm study and comparative analysis, separating the face from the image. When the correct face detection, localization using integral projection algorithm two eye centers, complete normalization of face image. As the location center points more accurately, the image preprocessing technology is appropriate, including image enhancement technology and the two values image processing technology, interference makes integral projection curve from the noise in a certain extent. KEY WORDS: Image Preprocessing, Face Detection, Identification of Human Eye, Color Space, Integral projection 目　录 前　言 1 第1章 绪 论 3 1.1 论文背景和意义 3 1.2 国内外发展现状和趋势 3 1.2.1 国内现状 3 1.2.2 国外现状 4 1.2.3 发展趋势 5 1.3 本文