(Excel 2016 VBA)
執行速度 快 慢
在Connecting 的過程中，可以發現當 Window 的中心點之間有一個空白 Pixel 的距離以上時，Connecting 就會發生做用，然而，在以往的設計中，都定位在 2 個 Window 中心點的中間或是偏往某一個 Window 中心點的方向。對於直線部份時是
無太大影響，但是對於斜線部份時，有時會放大失真的現象。 Window 中心點彼此靠得太近，造成 Connecting 後，在分岔處形成一條直線的問題。
所以可以設計一個往後退1~2 個 Window 中心點的功能來彌補參數 d 的值為 1 時，
所衍生出的此問題。
圖 6-1 Window 擴展故過大而包覆其它紋路示意圖
Window 擴展的大小，是一個很重要議題，由本研究的系統的模擬可以看出，
Window 要擴到恰到好處才能發揮分岔演算法的功能。Window 擴展的很小(參數 d
等於 1)，優點為可在紋路間距很小的情況下勝任；缺點為遇到紋路的分岔和匯合
範例圖案 Zhang and Suen[1]
Lu and
Wang[2] Baruch[3]
Chouinard
Microsoft Excel 2016 的崛起，代表微軟告訴我們要時時刻刻保持學習與創新。
Excel 2003 最大儲存格數只有 256 x 65536，這是無法催生出本研究的系統的，因 為隨便一個全指紋圖檔就超過256 了。直到 Excel 2007 出現後，大幅翻轉使用者 的習慣，最大的特點之一，是最大儲存格數擴增為16384x1048576，這一劇變，使 得本研究的系統的構想得以實現。
Line Following 的研究要有很多經驗與 trial and error(嘗試錯誤法)才能得知當 下的問題，而本研究的系統的圖像式動態展示，便成為了一大利器。不僅在圖像式 的除錯上或是檢視執行動態的過程中都具有很大的幫助。
網路的崛起與資訊的爆發，使得知識可以更快速傳遞，Line Following 的相關 文獻與相關文獻內的資料取得也是透過網路的便利。軟體一直在進步且時代一直 在變，細線化的演算法從侵蝕性的演算法，到運用人眼概念的Line Following 劃時 代啟發，由 Line Following 可以得知，其 Window 就代表了人類的眼睛，Window 能夠擴展的恰到好處，幾乎左右了Line Following 細線化結果的好壞，如果 Window 擴太大而導致 Contour Tracing 無法追蹤到 Window 內所包含的較短分支，這將發
生計算 Window 的邊界數與現實情況不符，這也是未來需要再做探討；反之，
Window 擴展的太小，也會阻礙了提前偵測出分岔的時機，容易形成細線化在分岔 點的失真。
最後，一個軟體或事物再怎麼優秀也只是在那個時代而已，如果沒有經常的 Update 或是隨著時代潮流的變化而做應變的話，遲早也是會走上沒落一途。如果 只是純研究的話就另當別論。畢竟現在的世界進步的很快很快，軟體的更新也得跟 上腳步，做為一個人類也得時常自我學習與進步，才不會被時代淘汰。
期待指紋的細線化能夠和AI(Artificial Intelligence)人工智慧做結合。之前的細 線化是運用一種連續侵蝕的方式，一直到Line Following 提出後，震撼整個 Thinning 學界，締造了一個里程碑。如同 AlphaGo 在與人類圍棋世界冠軍比賽中，令人驚
豔。希望未來能藉由AI 來執行細線化，這應該又會是下一個里程碑吧。
101
參考文獻
[1] T. Y. Zhang and C. Y. Suen, "A fast parallel algorithm for thinning digital patterns,"
Communications of the ACM, Volume 27, Issue 3, pp. 236-239, March 1984.
[2] H. E. Lu and P. S. P. Wang, "A comment on “a fast parallel algorithm for thinning digital patterns”," Communications of the ACM, Volume 29, Issue 3, pp. 239-242, March 1986.
[3] Orit Baruch, "Line thinning by line following," Pattern Recognition Letters, Volume 8, Issue 4, pp. 271-276, November 1988.
[4] Claude Chouinard and Rejean Plamondon, "Thinning and Segmenting Handwritten Characters by Line Following," Machine Vision and Applications, Volume 5, Issue 3, pp. 185-197, Summer 1992.
[5] E. S. Deutsch, "Thinning Algorithms on Rectangular，Hexagonal，and Triangular Arrays," Communications of the ACM, Volume 15, Issue 9, pp. 827-837, September 1972.
[6] O'Gorman L. (1988a) An analysis of feature detectability from curvature estimation.
Proceedings of Computer Vision and Pattern Recognition Conference pp. 235-240, Ann Harbor
[7] O'Gorman L. (1988b) Curvilinear feature detection from curvature estimation.
Proceeding of the 9th International Conference on Pattern Recognition pp. 1116-1119, Rome
[8] Abeer George Ghuneim, "Contour Tracing," Pattern Recognition course, Winter 2000,
[11] Wahyu Andhyka Kusuma and Lailatul Husniah, "Skeletonization Using Thinning Method for Human Motion System," 2015 International Seminar on Intelligent Technology and Its Applications, pp. 103-105, May 2015.
[12] Vicky Sintunata and Terumasa Aoki, "Grey-scale skeletonization using Delaunay
102
triangulation," 2017 IEEE International Conference on Consumer Electronics - Taiwan (ICCE-TW), pp. 57-58, June 2017.
[13] Vicky Sintunata and Terumasa Aoki, "Skeleton Extraction in Cluttered Image Based on Delaunay Triangulation," IEEE International Symposium on Multimedia, pp.
365-366, December 2016.
[14] Lynda Ben Boudaoud, Abderrahmane Sider, and Abdelkamel Tari, "A new thinning algorithm for binary images," 2015 3rd International Conference on Control, Engineering & Information Technology, pp. 1-6, May 2015.
[15] Wei Chen, Lichun Sui, Zhengchao Xu, and Yu Lang, "Improved Zhang-Suen thinning algorithm in binary line drawing applications," 2012 International Conference on Systems and Informatics, pp. 1947-1950, May 2012.
[16] Jianwei Dong, Wuhong Lin, and Chao Huang, "An improved parallel thinning algorithm," 2016 International Conference on Wavelet Analysis and Pattern Recognition, pp. 162-167, July 2016.
[17] Xu Zheng Qin, Li Lei, Qi Dong Yuan, Wang Qi, and Meng Ying, "An Algorithm of Thinning on Character Skeleton," 2013 International Conference on Computational and Information Sciences, pp. 730-733, June 2013.
[18] Kamlesh Tiwari, Gunakesh Tiwari, and Phalguni Gupta, "Extraction of high confidence minutiae points from fingerprint images," 2015 International Conference on Computer and Computational Sciences, pp. 238-243, January 2015.
[19] Iwasokun Gabriel Babatunde, Akinyokun Oluwole Charles, and Angaye Cleopas Officer, "Fingerprint matching by neighbourhood characteristics," 2013 Science and Information Conference, pp. 434-442, October 2013.
[20] Miloš V. Banđur, Brankica M. Popović, Andelija M. Raičević, and Dragan Ranđelović, "Improving minutiae extraction in fingerprint images through robust enhancement," 2013 21st Telecommunications Forum Telfor, pp. 506-509, November 2013.
[21] Amruta Pise and S. D. Ruikar, "Text detection and recognition in natural scene images," 2014 International Conference on Communication and Signal Processing, pp. 1068-1072, April 2014.
[22] Asem Othman and Arun Ross, "On Mixing Fingerprints," IEEE Transactions on Information Forensics and Security, Volume 8, Issue 1, pp. 260-267, January 2013.
[23] Rahul Kr Das, Abhishek De, Chandrajit Pal, and Amlan Chakrabarti, "DSP hardware design for fingerprint binarization and thinning on FPGA," Proceedings of The 2014 International Conference on Control, Instrumentation, Energy and Communication, pp. 544-549, February 2014.
[24] Ankit Shrivastava and Devesh Kumar Srivastava, "Fingerprint identification using
103
feature extraction: A survey," 2014 International Conference on Contemporary Computing and Informatics, pp. 522-525, November 2014.
[25] Wing-Nin Leung, C. M. Ng, and P. C. Yu, "Contour following parallel thinning for simple binary images," Systems, Man, and Cybernetics, 2000 IEEE International Conference, Volume 3, 8-11 October 2000.