【作者】 夏义勇；

【导师】 苑玮琦；

【作者基本信息】 沈阳工业大学 ， 计算机应用技术， 2004， 硕士

【摘要】 出于安全性和方便性的考虑，生物以别技术得到越来越多的重视，其中，指纹自动识别系统是当前最热门的应用，这是山于指纹采集方便，且具有唯一性。本文利用计算机和活体指纹录入仪，综合利用图像处理和模式识别技术，对活体指纹识别系统所应用的各种技术进行了探讨，描述了一个指纹识别系统的一整套实现方案，并提出了一些自己的观点。全文内容共分为六章：第一章为绪论；第二章为系统设计；第三章为指纹图像细节特征提取，包括指纹图像与背景分离、指纹图像增强和二值化、细线化、提取指纹细节特征点；第四章为指纹图像匹配，这里介绍了点模式匹配通常方法和对其的改进；第五章为本自动指纹识别系统的实验结果；第六章为结论部分。主要研究内容为：1 在指纹图像与背景分割部分，在采用方筹法后，利用相邻的分割结果信息确定指纹范围。2 在指纹增强和二值化部分，利用块方向图滤波的连接断线和分离粘连的特点，点方向图滤波对细节点精确的特点，把二者结合起来，使得二值化效果有明显改进。3 提取指纹特征点后，采用设定的几个规则去除伪特征点。4 在指纹图像匹配部分对点模式匹配方法进行了改进，使得匹配所需时间缩短。 本文较为系统地研究了自动指纹识别的理论和技术，在指纹图象方向信息提取和利用指纹图象方向信息提取指纹脊线方面的研究有所创新。本文可以为指纹技术人员的研究提供参考。更多还原

【Abstract】 More and more people pay attention to Biology identification technology for the security and convenience reason. Fingerprint recognition is the most important application of it. That is because that it is easy to get the fingerprint of someone you want and there is no same fingerprint in the world. Based on computer and live fingerprint scanner, using technology image processing and pattern recognition, this study investigates theory and methods of automatic fingerprint identification, implements algorithms of automatic fingerprint identification system(AFIS). The thesis is organized as follows. Chapter 1 is introduction. Chapter 2 is presents design of the automatic fingerprint identification system and related issues. Chapter 3 discusses fingerprint feature extraction and a new minutiae algorithm based on imperfect thinning fingerprint image. Chapter 4 deals with a minutiae matching algorithm based on point pattern matching. The last Chapter is conchision. The main contents are as follow:1 The method to segment fingerprint from background are studied. By selecting threshold which makes the number of pixels that are bigger than the threshold equal to the number of pixels that are smaller than it for local area. ,fingerprint image can be self-adaptively segmented from the background. After that we can make sure the final result through the information around the local area.2 The method of ridge extraction is studied. Using the orientation information of local area, masks for smoothing and enhancing are designed respectively. These masks can smooth the image along orientation of the ridges and enhance the image along orientation orthogonal to ridges at the same time. Results indicate that the method can connect breaks of the ridges and can dispart conglutinated ridges. As a result, the ridges that are single color are extracted correctly.3 After get the minutiaes from thinning ridges, erase some minutiaes that are not real minutiaes using several regulars.4 Improve the point pattern matching method to reduce the spending on time.This study systemically investigates theory and methods of automatic fingerprint identification, bring a new idea in acquiring minutiae extraction with accuracy.This study will be helpful for the people who relate with AFIS.更多还原