指紋識(shí)別算法研究
摘要
隨著計(jì)算機(jī)和網(wǎng)絡(luò)的迅速發(fā)展，人們對(duì)身份認(rèn)證的準(zhǔn)確性、安全性與實(shí)用性提出了更高的要求?；谏锾卣髯R(shí)別的智能身份認(rèn)證技術(shù)也逐漸受到廣泛的關(guān)注。在眾多的生物識(shí)別技術(shù)中，指紋識(shí)別技術(shù)是發(fā)展最早、應(yīng)用最廣泛的一種。指紋識(shí)別技術(shù)充分利用了指紋的普遍性、唯一性和永久性的生物特征，已逐步取代了傳統(tǒng)的基于標(biāo)志和數(shù)字的識(shí)別方式，目前在網(wǎng)絡(luò)、銀行、金融、醫(yī)療和安檢等行業(yè)均得到了廣泛應(yīng)用。本文對(duì)指紋識(shí)別系統(tǒng)的原理和基本過(guò)程進(jìn)行了分析研究，重點(diǎn)研究了指紋圖像預(yù)處理算法，并且進(jìn)行了驗(yàn)證。
在指紋圖像預(yù)處理部分，論文對(duì)預(yù)處理的各個(gè)步驟包括規(guī)格化、圖像分割、中值濾波、二值化、細(xì)化等以及各個(gè)步驟的方法進(jìn)行了深入的分析和研究，選擇了一種圖像預(yù)處理方案。
在指紋特征提取部分，采用基于Matlab實(shí)現(xiàn)的指紋細(xì)節(jié)特征提取方法，并給出了去偽算法。指紋特征提取是從細(xì)化后的指紋圖中得到細(xì)節(jié)特征點(diǎn)（即端點(diǎn)和分叉點(diǎn)），此特征點(diǎn)含有大量的偽特征，既耗時(shí)又影響匹配精度。采用了邊緣去偽和距離去偽，使得特征點(diǎn)去偽前后減小了近1/3，然后提取可靠特征點(diǎn)信息，以便實(shí)現(xiàn)指紋匹配。
在指紋匹配部分，本文采用基于細(xì)節(jié)點(diǎn)的指紋匹配算法，并進(jìn)行研究。

關(guān)鍵詞　指紋識(shí)別；預(yù)處理；特征提?。黄ヅ?br>Fingerprint Recognition Algorithm
Abstract
With the rapid Progress of computer science and network technique，An accurate，secure and practical techno1ogy of Personal identification becomes more and more important. Technology of Personal identification based on Biometrics has received extensive attention. Technology of Fingerprint recognition is the earliest one and is app1ied widely in the all techniques of biometrics recognition, then is taking full advantage of the universality, uniqueness and permanency of the fingerprint, and gradually has taken place of traditional identification method that is based on symbol and number. Nowadays the technology of fingerprint identification is fully used in network, bank, finance, insurance and security. This paper research the basic principles and process of the fingerprint identification system，and focus on the pre-processing algorithms of fingerprint image and finally verify. In the fingerprint image processing section, thesis on preprocessing steps including specifications, image segmentation, median filtering, binarization, refinement, and so on, and each step of the way to in-depth analysis and research, is an image processing program.
Part of the fingerprint feature extraction, fingerprint-based Matlab implementation details of feature extraction methods, and gives to the pseudo-algorithm. Fingerprint feature extraction is refined to get the details of fingerprint feature points (the endpoint and bifurcation points), this feature points contain a large number of false features, time-consuming and will affect the matching accuracy. Used and the distance to the edge of the false and counterfeit, makes the feature points to reduce the false front of nearly 1 / 3, and then extract a reliable feature point information, in order to achieve the fingerprint match.
Part of the fingerprint match, the paper-based fingerprint minutiae matching algorithm, and conduct research.

Keywords Fingerprint recognition; Pretreatment; Feature extraction; Matching
目錄
摘要…… I
Abstract II

第1章 緒論 5
1.1 指紋識(shí)別技術(shù)的背景 5
1.2 指紋識(shí)別研究的目的和意義 5
1.3 指紋識(shí)別在國(guó)內(nèi)外的研究現(xiàn)狀及分析 6
1.3.1 國(guó)外研究現(xiàn)狀 6
1.3.2 國(guó)內(nèi)研究現(xiàn)狀 7
1.3.3 研究現(xiàn)狀分析 7
1.4 本文研究的主要內(nèi)容及工作安排 8
1.4.1 研究的主要內(nèi)容 8
1.4.2 工作安排 8
第2章 指紋識(shí)別的組成及原理分析 9
2.1 指紋識(shí)別系統(tǒng)的組成 9
2.2 指紋識(shí)別的基本原理 9
2.2.1 指紋的基本特征 9
2.2.2 全局特征 9
2.2.3 局部特征 11
2.2.4 指紋識(shí)別的一般算法 13
2.3 本章小結(jié) 14
第3章 指紋圖像預(yù)處理 15
3.1 對(duì)指紋圖像進(jìn)行分割 15
3.2 對(duì)指紋圖像進(jìn)行二值化處理 16
3.3 對(duì)指紋圖像進(jìn)行細(xì)化處理 18
3.4 實(shí)驗(yàn)結(jié)果 21
3.5 本章小結(jié) 22
第4章 指紋圖像特征提取與匹配 23
4.1 指紋圖像特征提取 23
4.1.1 指紋圖像特征提取的方法 23
4.1.2 特征點(diǎn)的提取 23
4.1.3 指紋圖像特征的去偽 24
4.2 指紋圖像的匹配 26
4.2.1 指紋匹配算法綜述 26
4.2.2 指紋匹配過(guò)程 27
4.3 實(shí)驗(yàn)結(jié)果 28
4.4 本章小結(jié) 30
結(jié)論 31
致謝 32
參考文獻(xiàn) 33
附錄 34