自动指纹验证系统预处理与特征提取算法的研究与实现

【作者】 古清月；

【导师】 康慕宁；

【作者基本信息】 西北工业大学 ， 软件工程， 2007， 硕士

【摘要】 近年来，随着社会和经济的发展，人们对身份鉴别的准确性、安全性与实用性提出了更高的要求。指纹识别技术以方便易用、高准确率和低成本等诸多优势备受关注，已经成为使用最广泛的身份认证的有效手段，在电子商务、犯罪识别、信息安全等领域得到广泛的应用。 指纹是手指末端正面皮肤上凸凹不平的纹路。在人的生长早期形成后，终生保持不变。自动指纹识别系统是集光电技术、图像处理、计算机及网络、数据库技术、模式识别技术等于一体的综合性系统。针对自动指纹识别系统的运行过程，在总结该领域现有研究成果的基础上，本文对自动指纹验证系统中指纹图像预处理与特征提取环节的关键算法进行了研究，在PIV的计算机上用MATLAB7实现了目前本系统各个环节所使用的主流算法，并反复实验比较，实现了论文中所提到的算法。与传统系统设计相比，本系统进行了一些改进： 1、在指纹图像预处理过程中，将指纹有效区域分割与方向场计算相融合，并直接做出质量评测，大大缩短了处理时间，并且避免了对低质量指纹无效的处理。 2、在方向场计算后，使用一种自适应的图像增强算法，根据处理邻域是否存在分叉点分别使用不同的滤波器进行图像增强，取得了较好的增强效果。 3、在指纹特征提取阶段，使用一种基于二值化图像的特征提取算法，即能去除细化过程的处理时间，又能避免图像细化引入的伪特征，并且能够提取出足够真实、丰富的细节点，为进一步的特征匹配打下良好的基础。更多还原

【Abstract】 In recent years, as the development of social and economic, the more accuracy, safty, practicality of identification were required. Among all kinds of biometric identification technology, people pay a great deal of concern on fingerprint recognition technology for it’s many advantages, such as more convenient using, higher accuracy and lower cost. And as the most widely used and effective means of identity authentication, it have been widely applied in e-commerce, criminal identification, information security, and other fields.Fingerprint is a rugged finger skin lines. Once the human embryo created, the fingerprint will unchanged in whole life. Automated Fingerprint Identification System is an integrated system which concentrate the photoelectric technology, image processing, computer and networking, database technology, pattern recognition technology. In the view of the operation of the AFIS, based on the conclusion of the existing research results, I research the key algorithm in the fingerprint image preprocessing and feature extraction algorithm of the fingerprint authentication system in this paper. The author actualize the mainstream algorithm which used in the popular AFIS and experiment repeated, achieve the algorithm used in this paper with MATLAB7.Compared with the traditional system design, the system made some improvements:1. In the fingerprint image preprocessing course inosculate the step of the orientation field and the effective regional division, and evaluation the image quality directly. Using this algorithm can greatly reduce the processing time and avoid to handle of the invalid low-quality fingerprint.2. After estimate the fingerprint orientation field, an adaptive image enhancement algorithm which use different filters according to whether there have a furcation in the neighborhood was used to enhance the image and achieve a better effect.3. A feature extraction algorithm based on the binary image was used in the fingerprint feature extraction stage. Using this algorithm can not only get rid of the time used in thinning processing and avoid to introduce the pseudo-feature ,but also can extract the rich real minutia to make a sound foundation for further feature matching.更多还原