基于Hadoop的指纹识别系统设计

基于Hadoop的指纹识别系统设计(任务书,开题报告,论文12500字)
摘要
指纹识别已经是当今社会上较为热门的生物识别，因为指纹有唯一性，永不改变性的特征，所以指纹作为生物密码相比较于传统密码来说就更为方便快捷。但正因为指纹识别这些优点，所以指纹识别的适用范围十分的广泛，因此指纹库将会是一个庞大的数据库，如果需要进行指纹匹配操作，将花费大量时间。而当今社会又是一个信息大爆炸的时代，对于大量的信息数据处理，云计算又出现在人们的眼前。于是运用Hadoop中分布式文件系统（HDFS）和Mapreduce来处理数据。将指纹图像特征值提取出来，再利用Mapreduce并行处理，高吞吐量的特点来处理庞大的指纹数据库。
    JAVA是一种跨平台，适合于分布式计算环境的面向对象的编程语言[4-5]。本课题，是在win7系统下配置好hadoop环境，用ant制作出的hadoop插件，将插件添加到eclipse中建立hadoop开发环境。论文对MapReduce，HDFS技术，图像处理算法（灰度化算法，二值化算法，细化算法）分析介绍，运用opencv解决指纹图像特征点匹配。此系统可以输出对比出指纹图像特征点匹配度后图像，返回匹配指纹图像的特征值。
关键词：指纹识别；hadoop；指纹图像处理；Mapreduce

Fingerprint identification system based on hadoop
Abstract
Fingerprint recognition has become a popular biometric identification in today's society. Fingerprints are unique and never change their characteristics. Therefore,fingerprints are more convenient and faster than traditional passwords. However, due to the advantages of fingerprint recognition, the scope of application of fingerprint recognition is very wide, so the fingerprint database will be a huge database. If fingerprint matching operation is needed, it will take a lot of time. Today's society is also an era of big information explosion. For a large number of information and data processing, cloud computing appears before people.So use Hadoop Distributed File System (HDFS) and Mapreduce to process data. Extract the fingerprint image feature values, and then use Mapreduce parallel processing, high throughput features to handle a huge fingerprint database.
JAVA is a cross-platform, object-oriented programming language suitable for distributed computing environments. This topic is to configure the Hadoop environment under win7 system, create Hadoop plug-in with ant, and add the Hadoop plug-in to eclipse to create Hadoop development environment. The paper introduces and analyzes the MapReduce, HDFS, image processing algorithms (gray leveling algorithm, binarization algorithm and refinement algorithm), and uses opencv to solve fingerprint point feature matching. The system can output the image after comparing the matching degree of the fingerprint image feature points and return the feature value of the matching fingerprint image.
Keywords:Fingerprint recognition;Hadoop;Mapreduce

目录
摘要    I
Abstract    II
第一章    绪论    1
1.1课题背景及意义    1
1.2国内外研究现状    2
1.3课题主要研究内容    3
1.4论文结构    4
第二章 课题相关技术    5
2.1Hadoop分布式计算平台    5
2.1.1Mapreduce    5
2.1.2HDFS    6
2.2Eclipse上Hadoop环境搭建    6
第三章设计方案    9
3.1指纹图像预处理    9
3.1.1灰度化处理    9
3.1.2二值化处理    12
3.1.3细化处理    13
3.2特征值提取以及比对    15
3.3用HDFS和Mapreduce处理特征值文本    21
第四章程序运行测试    27
4.1图像预处理阶段测试    27
4.1.1灰度化测试    27
4.1.2二值化测试    27
4.1.3细化测试    28
4.2图像特征点对比测试及提取特征点测试    28
第五章总结与展望    31
5.1本文主要工作    31
5.2进一步的展望    31
参考文献    33
致谢    35