指紋細節(jié)點提取與去偽算法研究


頁數(shù)：42 字數(shù)：20570


摘 要

步入信息化時代后，電子商務(wù)、自動銀行等業(yè)務(wù)逐漸開展。保障這些業(yè)務(wù)順利進行的關(guān)鍵之一為準確的身份鑒定。生物鑒定，即利用人們的生理特征或行為特征進行鑒定，比其他密碼、身份證等鑒定方式能更可靠、更可行的鑒定授權(quán)用戶及冒名頂替者。
指紋是人終生不變的特征之一，而且科學研究發(fā)現(xiàn)不同的手指指紋不同，即便同卵雙生的人的指紋也是如此。相比其他的生物鑒定方法，如簽名、臉部識別、聲音等，指紋鑒定具有費用低，效果好的特點，而且要可靠的多。除了在鑒定疑犯等警務(wù)司法工作中使用之外，指紋鑒定在人們的日常生活中也有廣泛應用。比如電子商務(wù)等?？梢赃@樣說指紋是人們隨身攜帶的身份證。
目前市場上流行的自動指紋身份識別系統(tǒng)主要由指紋特征的提取和匹配兩大模塊組成，其中指紋特征的提取又包括對指紋圖的濾波增強，特征的提取和虛假特征的祛除。本文主要研究指紋特征的提取和虛假特征祛除方面的算法。我們首先對生物特征鑒定技術(shù)，包括指紋識別技術(shù)做一簡單的介紹，然后就目前國內(nèi)外指紋特征提取，去偽的現(xiàn)狀給與概述，最后詳細描述我們所提出一套完整有效的指紋特征提取并除去偽節(jié)點的算法和實驗結(jié)果。

關(guān)鍵詞：生物特征鑒定技術(shù)，指紋識別，預處理，細節(jié)提取，匹配，終止點，分叉點，去偽
Abstract

In the information age, electronic commerce and automated banking become more and more popular. For the safety of these business accurate automatic personal identification is critical. Biometrics, which refers identification base on physical features or behavior features, compared with other identification method such as password, person identification card, is more reliable and more capable to identify impostor from authority user.
Fingerprint is an immutable feature and scientists find different finger has different fingerprint, even the twins have different fingerprints. Unlike other biometric such as signature, face, speech etc. fingerprint identification is much more reliable and more effective. Besides using in criminal identification and police work, it has become popular in civilian applications, such as electronic commerce. In other word fingerprint is a sort of identity card that people carry with them continuously.
The most Automatic Fingerprint Identification System are mainly consisted of two parts: Features Extraction and Minutia Matching, and the enhancement of fingerprint image, the extraction and remove spurious minutia can be included in the Features Extraction . This paper concentrate on the algorithm of fingerprint minutia extraction and purification, we firstly introduce the biometrics technique and the background of fingerprint identification, then summarize the popular extraction and purification algorithm domestic and abroad, finally we will give a particular presentation of our algorithm in fingerprint features’ extraction and purification and analyse the experiments result.

Key words: Biometrics, Fingerprint identification, preprocess, feature extraction, matching, Termination, Bifurcation, Feature Purification

目 錄
摘 要 2
Abstract 3
1.緒論 6
1.1．生物鑒定技術(shù) 6
1.1.1生物鑒定特征的要求及分類 6
1.1.2生物鑒定系統(tǒng)及工作模式 7
1.1.3不同生物鑒定技術(shù)的比較 8
1.2生物鑒定技術(shù)的應用 9
1.3 本文選題和結(jié)構(gòu) 9
本章小結(jié)： 10
2．指紋識別技術(shù) 10
2．1指紋鑒定的歷史 11
2．2指紋的采集 12
2．3指紋的分類 13
2．4指紋的特征描述及匹配 14
2．5自動指紋身份鑒定系統(tǒng) 16
2．5．1系統(tǒng)的模塊組成 16
2．5．2 系統(tǒng)的性能指標 17
本章小結(jié): 18
3．指紋特征的提取 19
3.1在細化圖上的提取算法 19
3.2 直接灰度提取 22
本章小結(jié) 24
4．本文實現(xiàn)的提取算法和實驗結(jié)果 24
4.1 指紋圖的預處理 24
4.2 細化 26
4.3 提取細節(jié)點 28
4.4 去除偽節(jié)點 29
4.4.1二值圖增強 30
4.4.2除去邊界效應的偽節(jié)點 31
4.4.3 畫窗標記去偽 32
4.4.4去除斷點 34
4.5實驗結(jié)果 36
五．全文總結(jié) 37
六．致謝 38
參考文獻 39