具重疊指紋及汙點雜訊之指紋圖像辨識方法探討

近年來，自動辨識指紋在資訊安全應用與刑事偵查上日益顯得重要。然而，目前指紋辨識的研究大多只針對單枚的指紋圖像進行辨識，探討如何增加辨識精確度與提升執行速度，但卻未考慮指紋圖像中可能含有其他枚指紋或污點雜訊的情況，並不符合真實的應用。因此，本論文嘗試探討指紋重疊的辨識問題。假設一給定的灰階圖像中包含了一枚完整的主要指紋與另一局部的指紋，此局部的指紋與主要指紋彼此重疊若干區域，但強度可能相等或較弱於主要指紋，我們將其稱之為「干擾指紋」。此狀況存在於偵查現場所採集的指紋可能受另一枚指紋所覆蓋，或在安全系統中，感測器因經過重複使用後未擦拭清除其上的指紋壓印，造成掃瞄時出現多枚不同的指紋。經實驗測試，現行的指紋辨識方法對於重疊指紋圖像的辨識能力趨近於零。為了使重疊指紋圖像得以辨識，本論文提出一種分割與消除干擾指紋的方法。實驗結果證實在干擾指紋強度較主要指紋為弱的情況下，本系統能有效地辨識出主要指紋。

關鍵字

汙點雜訊；指紋重疊；指紋強度

並列摘要

Automatic identification of fingerprint images is gaining importance as the increasing demands in information security and criminal investigation. However, existing researches do not fully tackle the fingerprinting problem under adverse conditions, such as noise or overlapping fingerprints, but only focus on improving the identification accuracy and efficiency. In this work, we discuss a more practical situation that the fingerprint to be identified is corrupted by block noise or another overlapping fingerprint. It is assumed that a given gray-scale fingerprint image contains a “primary fingerprint” overlapped by an “interfering fingerprint”. The interfering fingerprint is only part of a fingerprint with intensity equal to or lower than that of the primary fingerprint. Such an overlapping case may appear in real criminal scenes or security systems whose fingerprint sensor is not cleaned after a while of use. Our experiments found that start-of-the-art approach cannot handle the problem of identifying fingerprints with block noise or overlapping fingerprints. To solve this problem, this work proposes some methods to locate and separate the interferences from the primary fingerprint. The experiments demonstrate a significant improvement achieved with the proposed methods.

並列關鍵字

Fingerprint ； dentification ； Overlapping fingerprints ； Block noise

參考文獻

[1]. 曾銘崧，熱感式指紋成像及其自動辨識法之研究，碩士論文，私立中原大學電子工程系，桃園，2001。

[3]. H. Lin, Y. Wan and A. Jain, “Fingerprint Image Enhancement:Algorithm and Performance Evaluation,” IEEE Trans. Pattern Analysis and Machine Intellignce, vol. 20, no. 8, pp. 777-789, Aug. 1998.

[4]. Y. Wang, J. Hu and D. Phillips, “A Fingerprint Orientation Model Based on 2D Fourier Expansion (FOMFE) and Its Application to Singular-Point Detection and Fingerprint Indexing,” IEEE Trans. Pattern Analysis and Machine Intellignce, vol. 29, no. 4, Apr. 2007.

[5]. X. Peng, G.. Wu, Z. Peng and Y. Li, “An Improved Method of Fingerprint Singular Points Detection,” in Proc. Int. Conf. Control and Automation, Guangzhou, China, May, 2007.

[7]. J. S. Weszka and A. Rosenfeld, “Histogram Modification for Threshold Selection,” IEEE Trans. Systems, Man, and Cybernetics, SMC- 9, pp. 38-52, 1979.

國際替代計量

具重疊指紋及汙點雜訊之指紋圖像辨識方法探討

全文下載

主題瀏覽