【作者】 江虹；

【导师】 郭树旭；

【作者基本信息】 吉林大学 ， 电路与系统， 2011， 博士

【摘要】 为了解决基于手部特征的识别技术实用化进程中存在的问题,本文对基于多层次、多特征的手部生物特征图像进行分析来实现人身份的鉴别,并对基于图像内容的手部生物特征的识别技术进行了较为深入的研究。本文主要研究了基于手部生物特征的识别技术中包括多指指纹图像获取技术,多指指纹图像有效区域的分割,手部血管图像的采集技术,血管特征提取算法,纹形分类技术等关键问题的研究。本文首先探讨了手部生物特征的特点,并对基于手部生物特征的识别系统及关键技术进行详细的归纳总结。其次,为了解决海量指纹数据库检索耗时大、复杂性高和不精确的问题,提出基于指纹奇异点结构的纹形分类方法。再次,本文给出了基于光学技术的多指指纹采集方案及实现方法,以及基于光学技术的人体手足末端血管采集的实现；最后研究并提出了多尺度曲率下的血管图像ROI自动检测定位的新算法,实现ROI的精准定位。更多还原

【Abstract】 Biometrics is the measurement and analysis of biological data, science and technology, is used to measure, statistics and analysis of biological data, a discipline and technology, using human physiological or behavioral characteristics to identify the unique user identity. Read in full on the basis of large number of documents, the paper covered areas of research the status of the analysis and research later, the Ministry for human exposure to hand-based biometric identification technology, with the current trends in research and development, this Department starting from a variety of Hand features biometric collection technology, working on hand a number of key biometric technologies, including image acquisition based on optical technology, technical studies, access to image quality assessment, biological feature extraction and matching technology, combined with information and sensor fusion theory and hands Department of feature recognition applications, to achieve technological innovation. For more content, the organization of this paper consists of five parts:The first part is the summary of this thesis, and it describes the purpose of this paper and the subject of the source, and then describes the identity-based biometric authentication technology researches and the problems are still in existence. In the finally, an overview of the main contents of this article, by the introduction of biometric identification technology background, significance, research and application at home and abroad to start on the dissertation research and explain the contents of arrangements;The second part of this paper, which is hand-based biometric identification technology, is the information theory and the test database. Hand biometric image is part of the whole system is based on the sample set to establish inspection and evaluation system has played an important role, helping the follow-up process.The third part of the optical system based on fingerprint technology and multi-fingered hand of human vascular optical image capture technology. Hand biometric image acquisition is an important part of the whole system, collecting device is a recognition algorithm based on a direct impact on image quality, image acquisition device for quality assessment needed to help the follow-up process. This section is proposed and implemented an optical system of the vascular image acquisition technology, and on this basis builds a biometric database of hand features;The fourth part of the hand-based biometric authentication system, the key issue. Fingerprint recognition technology in terms of the theory of algorithms or from the engineering point of view, and the research results are rich. Fingerprint feature extraction, feature points up than the other already has a very mature technology, but for large networks based on the system than the recognition speed to be improved, this article from the biometric identification system of hand the key problems start, in Research on the search strategy, given the large search strategy than the system. This section also presented in this chapter the multi-scale curvature analysis method to obtain blood vessel model, focused on the study of vascular pattern extraction algorithm, the angle of curvature analysis of multiple blood vessels extraction issues.The fifth chapter is experimental and conclusions of each chapter, and the prospects for future work.更多还原