基于三维鼻形的生物特征识别-通信与信息系统专业论文.docxVIP

摘 要 鼻形识别是一种新兴的生物识别方法，近年来受到了越来越多国内外研究者 的关注。和现有的生物特征识别方法相比，鼻子最重要的特征是不易改变性，稳 定性，难以隐藏性等，并且能够采取非接触的方式进行采集，不需要识别对象较 多的配合，特别适用于秘密监视和跟踪的场合。为了把握它稳定丰富的空间结构 信息，使用三维的方式最有利于对鼻子进行检测和获取。在三维鼻形识别方面， 各种理论和方法都不太完善，还处于初级阶段。 本论文的研究内容如下： 1.阐述了鼻形识别的研究背景和意义以及国内外发展现状，并重点介绍了三 维鼻形识别。 2.本文提出了一种定位鼻子区域特征点的新方法，通过分析鼻子的结构特征， 提取鼻子区域曲线，然后分析曲线形状，从而实现鼻子区域鼻尖点，鼻根点，鼻 下点，左右侧鼻翼点这五个特征点的自动标定。实验表明，该标定算法有较快的 速度，并对表情，姿态，光照的变化有较强的鲁棒性。 3.本文提出基于几何特征的三维鼻形识别算法，充分考虑鼻子的三维几何特 征，首次将鼻子区域的角度特征，体积特征加入到几何特征中，将鼻形用一组几 何特征向量来表示，最终将识别归结为特征向量之间的匹配。 关键词： 三维人脸 三维鼻形识别 特征点 几何特征 加权最小距离分类器 Abstract As an emerging biometric identification method, the nose shape recognition has been paid more and more attention by researchers at home and abroad. Compared with existing biometric identification methods, the most important characteristic of nose is not easy to change, stability, and difficult to hide. And it can be collected by a non-contact way, not need more cooperation of identify objects, particularly suitable for secret surveillance and tracking occasion. In order to grasp its stable and rich spatial structure information, three-dimensional manner is helpful for nose detecting and obtaining. The three-dimensional nose shape recognition’s theories and methods are not perfect, still in the primary stage. The main research contents in this paper are as follows: The research background, significance and the current research situation of the nose shape recognition are illustrated. And put emphasis up on introducing the three-dimensional nose shape recognition. A new method of positioning feature points of nose region is proposed. Through the analysis of the structure characteristics of nose, extraction nose area curves and analysis the shape of curves, finally realize the five feature points’ automatic calibration, including nose tip, nasion, subnasale, alar base. Experiments show that the calibration algorithm has faster speed, for expression ， posture and illumination changes have strong robustness. The three-dim