現今應用最為廣泛的生物辨識技術為指紋辨識。國際標準組織(ISO)已於2005年完成制定並且公告指紋特徵檔標準格式(ISO19794-2)。指紋特徵點定義為指紋紋路的端點與分歧點兩種類型。而指紋特徵點的訊息包含了特徵點的座標、方向與型態。

指紋比對在應用上常需進行大量指紋資料的比對，如果使用一般的特徵點比對方法將會較花時間，增加整體的比對時間。目前我們採用的指紋辨識系統利用三角形比對迅速篩選出可能匹配的指紋，再進行特徵點比對，選出特徵點比對分數最高的指紋，即為匹配指紋。不過特徵點品質的好壞，會間接造成比對上的影響。指紋影像容易因空氣溼度與手指按壓的影響，即便是同一隻手指頭，每次取得的指紋影像不可能完全相同，使得指紋特徵點的數量、座標與方向亦不盡相同。品質不好的指紋影像造成特徵抽取產生兩特徵點距離過近的不穩定特徵點，在三角形比對中的三角形，是依據特徵點的分佈產生，所以往往不穩定特徵點因距離過近而產生出狹窄的三角形，這些狹窄的三角形在比對過程是無法與其他三角形所匹配。

本篇論文藉由增加穩定的三角形，來提升三角形比對辨識率，使系統在三角形比對後，匹配的指紋能夠落在我們選取指紋數量內的機率提高，在特徵點比對時，找出匹配的指紋機率也將提高，進而提升系統的辨識率。由於三角形比對的時間遠小於特徵點比對時間，所以我們能在系統辨識率幾乎不變的情況下，減少三角形比對後選取指紋的數量，降低特徵點比對次數，使系統比對速度加快。

另外，嵌入式系統應用越來越熱門，相信未來指紋辨識系統也不會只有在傳統的PC上發展，所以我們希望將設計後的指紋辨識系統移植至ARM平台，期望指紋辨識在應用上能更加廣泛。嵌入式處理器有許多種，像是：ARM、PowerPC、MIPS、SC-400…等，選用ARM平台開發指紋辨識系統主要有三大優勢：市占率最高 、效能最好 耗電最低，透過此優勢開發出實用的指紋模組。而系統在ARM平台執行，三角形比對的時間遠小於特徵點比對時間會更加明顯，故辨識速度之效果也提升許多。

本論文採用FVC2002四組指紋資料庫。根據實驗結果，在系統辨識率幾乎不變的情況下，能夠使三角形比對後篩選資料庫1%的指紋數量，減少至0.75%，使特徵點比對的次數減少，提升系統比對速度。整體系統比對速度在PC平台上提升11.19%至21.58%的效果，在ARM平台上，因三角形比對的時間與特徵點比對時間的差距更加明顯，提升系統比對速度效果達22.3%至27.35%。

The fingerprint recognition is the most widely used biometric recognition technology nowadays. International Organization for Standardization (ISO) had established and announced the standard format for the fingerprint template (ISO19794-2) since 2005.The minutiae is defined as ridge ending and ridge bifurcation, and minutiae information contains coordinate, direction and type.

The matching speed under large scale fingerprint database is a challenge of fingerprint identification application. The minutia-based fingerprint matching method takes more matching time in general. Therefore, we use the triangular match quickly filter out the possibly correct fingerprint, and then use minutiae to find the correct fingerprint in the match. The quality of minutiae impacts fingerprint match indirectly. The fingerprint images are easily effected by air humidity and finger press. Even the same fingers, the fingerprint image can’t be obtained exactly the same each time, so the number of the minutiae, coordinates and the direction of the minutiae are different. The too closed unstable minutiae are produced by bad quality fingerprint images. The triangle is generated by minutiae, so unstable minutiae will generate narrow triangle. These narrow triangles can’t be match with other triangles in the fingerprint match.

In my thesis, we would like to improve the recognition rate of triangle match by increasing stable triangles. The system after the triangular match process, the correction rate of selected fingerprints will increase. Because the triangle matching time is much shorter than minutiae matching time, thus we can decrease system total time under same system recognition rate by reduce the number of minutiae matching.

On the other hand, the embedded system is getting more and more popular. We believe the fingerprint recognition system will not only develop in traditional PC. Hence we porting the designed fingerprint recognition system to the ARM platform, so that the usage of fingerprint recognition will increase. There are several kinds of embedded processor, such as arm, PowerPC, MIPs, SC-400 and so on. The main reason to choose the ARM platform to develop fingerprint recognition system as follows: The market share is the highest, it is the most effective and consume least electricity. We could develop practical fingerprint module from these advantages. If the system runs on ARM platform, the time of triangle match will be much lesser than minutiae match. Hence the recognition speed will also be fasten.

This research adopts four sets of the FVC2002 fingerprint database. According to the experiment result, while the system recognition rate remain the same, we could select 1% of fingerprints out of the database after triangle match. It decreases to 0.75% and the number of times for minutiae match also decreases, while the matching speed became faster. The system speed increase between 11.19% and 21.58% in PC platform, between 22.3% and 27.35% in ARM platform.

摘要 i
英文摘要 iii
致謝辭 v
目錄 vi
圖目次 viii
表目次 x
第一章 緒論 1
1.1 生物辨識 1
1.2 指紋辨識系統 6
1.3 研究動機與目標 10
1.4 論文架構 11
第二章 相關研究 12
2.1 指紋識別之方法 12
2.2 指紋模組之硬體架構 17
第三章 指紋識別方法之研究 19
3.1 系統架構 21
3.2 指紋特徵點之定義 23
3.3 三角特徵之產生 24
3.3.1 刪除相近之特徵點 24
3.3.2 Delaunay triangulation 25
3.3.3 三角特徵之定義 28
3.3.4 三角形集1與三角形集2之合併 30
3.4 三角特徵比對流程 32
3.5 特徵點比對程序 34
第四章 系統實作與實驗結果 38
4.1 系統實作 38
4.1.1 指紋辨識系統介面 40
4.1.2 RS232傳輸指令 42
4.1.3 ID_List與User_Data 配置 44
4.2 指紋資料庫 47
4.3 實驗結果 49
4.3.1 系統錯誤拒絕率 49
4.3.2 系統比對時間效果 59
4.4 實驗結果探討 62
第五章 結論與未來研究方向 63
參考文獻 65

[1]D. Maltoni, D. Maio, A. K. Jain, and S. Prabhakar, “Handbook of fingerprint recognition”,2st ed. Springer Verlag London Limited, 2009.
[2]A. K. Jain, “Who’s who ? Challenges in biometric authentication,” in DAGM Symposium, pp.33-35,2003.
[3]International Biometric Group “Biometrics Market and Industry Report 2009-2014”, http://www.ibgweb.com/ .
[4]Moenssens, Fingerprint Techniques. London: Chilton Book Company, 1971.
[5]J. D. Huang and W. H. Hsu, “Quality analysis and study of minutiae in fingerprint images,” pp. 504-530,2006.
[6]Ning Liu, Yilong Yin, and Hongwei Zhang, “A Fingerprint Matching Algorithm Based On Delaunay Triangulation Net,” in The Fifth International Conference on Computer and Information Technology, pp.591-595,2005.
[7]Y. Fumio, I. Seigo, and E. Shin, “Real-time fingerprint sensor using a hologram,” Applied Optics, vol. 31, no. 11, pp. 1794-1802, 1992.
[8]L. Hong, Y. Wan, and A.K. Jain, “Fingerprint Image Enhancement: Algorithm and Performance Evaluation,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 20, no. 8, 1998.
[9]A. Jain, L. Hong, and R. Bolle, “On-line fingerprint verification,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 19, no. 4, pp.302-314, 1997.
[10]P. Baldi, and Y. Chauvin, “Neural networks for fingerprint recognition,” Neural Computation, vol. 5, pp.402-418,1993.
[11]B. Mehtre, “Fingerprint image analysis for automatic identification,” Machine Vision and Applications, vol. 6, pp.124-139, 1993.
[12]N. Ratha, K. Karu, S. Chen, and A. Jain, “A real-time system for large fingerprint databases”, IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 18,no. 8, pp.799-813,1996.
[13]J. L. Blue, G. T. Candela, P. J. Grother, R. Chellapa, and C. L.Wilson, “Evaluation of pattern classifiers for fingerprint and OCR applications,” Pattern Recognition, vol. 27, no. 4, pp. 485-501, 1994.
[14]M. Kawagoe and A. Tojo, “Fingerprint pattern classification,” Pattern Recognition, vol. 17, pp. 295-303, 1984.
[15]K. Karu, and A.K. Jain , “Fingerprint classification,” Pattern Recognition, vol. 29,no. 3, pp. 389-404, 1996.
[16]D. Maio, and D. Maltoni, “A Structural Approach to Fingerprint
Classification,” in Proc. Int. Conf. on Pattern Recognition (13th), 1996.
[17]Jain Anil, Prabhakar Salil, and Hong Lin, “A Multichannel Approach to Fingerprint Classification,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 21, pp. 348-359, 1999.
[18]R. Germain, A. Califano, and S. Colville,“ Fingerprint matching using transformation parameters, ”IEEE Computational Science and Engineering, vol. 4, no. 4, pp.42-49, 1997.
[19]G. Bebis, T. Deaconu, and M. Georgiopoulos,“ Fingerprint Identification Using Delaunay Triangulation,” IEEE International Conference on Information Intelligence and Systems, pp. 452-459,1999.
[20]A. A. Saleh, and R. R. Adhami, “Curvature-based matching approach for
automatic fingerprint identification,” IEEE Proceedings of the 33rdSoutheastern Symposium on System Theory, pp. 171-175, March2001.
[21]B. Bhanu, and X. Tan, “Fingerprint indexing based on novel features of minutiae triplets,” IEEE Transactions on Pattern Analysis and Machine Intelligence, pp.616-622, May 2003.
[22]J. P. Starink, and E. Backer, “Finding point correspondence using simulated annealing,” Pattern Recognition,pp.23 l-240,1995.
[23]A . Ranade, and A . Rosenfeld,“ Point pattern matching by relaxation” ,Pattern Recognition,pp.269-275,1993.
[24]Jinhai Zhang , “Research on embedded fingerprint identification system”, Second International Conference onMechanic Automation and ControlEngineering ,pp.594-597,2011.
[25]Yanpeng Wang, Qing Li, and Li Zhang, “Design of Embedded Fingerprint
Identification System Based on DSP”, IEEE International Conference onAnti-Counterfeiting, Security and Identification,pp.47-50,2011.
[26]Zhang Jinhai , “Fingerprint identification system based on DSP”, Technology Conference on Cross Strait Quad-Regional Radio Science andWireless,pp.1071-1074,2011.
[27]S.H. Chang, F.H. Cheng, W.H. Hsu, and G.Z. Wu, “Fast algorithm for point pattern matching :Invariant to translations, rotations and scale changes,” Pattern Recognition, vol. 30, pp.311-320,1997.
[28]X. Liang ,A. Bishnu, and T. Asano, “A Robust Fingerprint Indexing Scheme Using Minutia Neighborhood Structure and Low-Order Delaunay Triangles,” Information Forensics and Security,pp.721-733,2007.
[29]The FVC2002 Database, http://bias.csr.unibo.it/fvc2002.