资源预览内容
第1页 / 共9页
第2页 / 共9页
第3页 / 共9页
第4页 / 共9页
第5页 / 共9页
第6页 / 共9页
第7页 / 共9页
第8页 / 共9页
第9页 / 共9页
亲,该文档总共9页全部预览完了,如果喜欢就下载吧!
资源描述
.基于几何特征的人脸识别学号:姓 名:(XX大学 机电工程与自动化学院,200072)摘要:人脸识别技术作为多学科领域的、具有挑战性的课题,它覆盖了数字图像处理、模式识别、神经网络、数学等诸多学科的内容,同时也具有十分广泛的应用价值。在人脸识别领域,基于几何特征的人脸识别算法因其计算简单、使用有效等特性,引起了人们的广泛注意,并已成为人脸图像特征提取和识别的主流方法之一。本文定位人脸器官,通过人脸面部拓扑构造几何关系的先验知识,利用基于灰度投影的方法在知识的层次上提取人脸面部主要器官特征,将人脸特征用一组几何特征向量表示,识别归结为特征向量之间的匹配。本文工作包括:(1)对灰度积分投影理论进展了详细的介绍和分析。这种方法是目前定位人脸轮廓的主要方法。在此根底上对一种新的在竖直方向上定位人脸左右轮廓的灰度差投影法进展了改良。投影法本质上是一种基于统计的方法,在具体应用时又结合了人脸特征分布的先验知识。这种方法不需要对积分投影图做任何平滑处理等操作,因而算法简单,准确率高,速度很快。(2)提出了一种准确定位眼睛的方法,该算法将眼区灰度总体分布特点与眼部灰度变化特点相结合,将传统的积分投影法与灰度差累加值投影法相结合,通过大量试验选取适宜的参数。试验结果说明,该算法对光照变化不敏感,定位准确率高。运用灰度积分投影结合人脸特征的先验知识定位鼻子,这种定位方法得到的准确率也是比拟高的。嘴巴的定位那么利用投影法求得。(3)本文根据定位出来的人脸器官选出七个特征点,即四个眼角点、鼻尖点和两个嘴角点。利用它们构造了十个特征向量,并对其进展归一化运算。图像识别的最后一个过程就是分类,本文采取根据模式相似性的最近距离分类器进展分类。用加权比值函数来计算特征相似度,更适用于人脸图像的识别和计算。如何选择出适宜的识别门限是个很复杂的问题,还有待于进一步研究。关键词:人脸识别;灰度投影;几何特征;特征提取The Research of Face Recognition Algorithm Based on GeometricFeaturesStudent number:15721637 Name:Zhao Pei-pei(Institute of electrical and mechanical engineering and automation, Shanghai University, Shanghai 200072, China)Abstract:the technology of human face recognition as a multi-disciplinary field and challengingpsubject which contains digital image processing,pattern recognition,puter vision,neuralnetwork,psychology,physiology,mathematics and a good many subjectsIn the meantime,italso has widely usedIn the field of face recognition,the method of human face recognitionbased on geometric features has been paid great attention for its simple calculation andavailabilityAt present,it has bee one of the dominant methods as the feature extractionand recognitionTMs article locates human face organs,through apriori knowledge of humanface topological structure geometrical relationship,making use of method based on constructto extract the features of human face organs,expressing human face through a set ofgeometric feature vectorsThe recognition putting in summary is matched with feature vectorThis paper includes the following parts:(1)Have a detailed introduction and analysis about the theory of greyscale integratedprojectionThis method is now the main method of locating human faceWe put forward anew method called greyscale differential projection which is based on the previous methodand locating the contour of human face vertical directlyProjection method is essentiallybased on statisticsIt bines the apriori knowledge of human face feature distribution inthe applicationThis method neednt to do any pretreatment to the image and any smoothingtreatment to the integrated projection imageSo this algorithm is simple;the accuracy is high; the speed is quick(2)Give an introduction about the method of locating eyes preciselyThis algorithmbines the character of the eye area greyscale totally distribution and greyscaletransformation;bines the methods of traditional integrated projection and differentialprojectionThe experiment led to the fact that this algorithm is not sensitive to theillumination transformation and has a high accuracyUsing greyscale integrated projectionbines the apriori knowledge of human face character to locate noseThis location methodalso has high accuracyThe location of mouth is abtained through projection method(3)The choice of characteristic points needs enough information and cant go so far as toincrease calculation quantityThis article chooses seven characteristic points,namely,fourcanthus points,tip of nose and two corners of mouth pointsConstruct ten eigenvectors usingthem and carries on the normalization calculation to themThe last process of imagerecognition is classificationAfter adopting some standards to extract feature of humanimages,we construct category separability decision rule according to these characters anddesign classifierThis article takes use of minimum distance classification to classifyIt iSmore suitable for human face recognition and calculation using weighing ratio to calculatesimilarityHow to choose a suitable recognition threshold is a difficult problem and needfurther researchThis article ascertains it through a good many experimentsKey Words:Face Recognition;Greyscale Projection;Geometric Characters;FeatureExtraction1. 引言1.1人脸识别技术:人脸识别是一个活泼的研究领域,是人类视觉最出色的能力之一。虽然人脸识别的准确性要低于虹膜、指纹的识别,但由于它的无侵害性和对用户最自然、最直观的方式,使人脸识别成为最容易被承受的生物特征识别方式。应用领域:人脸识别系统在金融、证券、社保、公安、军队及其他需要平安认证的行业和部门有着广泛的应用。典型应用:1罪犯调查 3重用门票2访问控制 4信用卡 1.2人体生物认证技术人脸识别是人体生物认证技术的一种,人体生物的生物特征包括生理特征和行为特征两大类。 人体的生理特征主要包括人脸、指纹、掌纹、掌形、虹膜、视网膜、静脉、DNA、颅骨等,这些特征是与生俱来的,是先天形成的; 而行为特征包括声纹、签名、步态、耳形、按键节奏、身体气味等,这些特征是由后天的生活环境和生活习惯决定的。这些生物特征本身固有的特点决定了其在生物认证中所起的作用是不同的。生物特征识别:人脸 脸部热量图 指纹 签名 Face Face heat figure Fingerprint Signature图 1Fig.1常用生物特征的比拟:生物特征普遍性独特性稳定性可采集性性能承受程度防欺骗性人脸高低中高低高低指纹中高高中高中高手形中中中高中中中虹膜
网站客服QQ:2055934822
金锄头文库版权所有
经营许可证:蜀ICP备13022795号 | 川公网安备 51140202000112号