在電腦視覺與影像研究的領域上，距離轉換和分水嶺的搭配是一項重要的技術，利用分水嶺轉換的計算，可萃取出影像的重要特性，如物體輪廓(object boundary)和數量辨識等應用上。
在此篇論文中，將比較我們所提出的分水嶺轉換方法與其他的分水嶺轉換方法的效果，我們提出的方法將更準確。在影像處理中，物體的邊界分割一直是一個重要且基本的課題。本研究改良傳統式的分水嶺轉換，提出一個選舉式的概念，使相互接觸之物體影像可以正確分割。藉由此方法處理之物體影像，可以依據其歐式距離轉換的結果，很容易求得影像中各物體之正確輪廓，方便在醫學影像上的應用。最後與National Institutes of Health(NIH) 所開發的圖像處理軟體ImageJ以及後來其他學者提出的控制標記分水嶺轉換來進行比較。

Euclidean Distance transformation and watershed are fundamental technique for the application fields of image understanding and computer vision. Calculated using the watershed transformation can be extracted important image features. Such as the identification of the contour and the number of such applications.
In this paper, we will compare the effects of our proposed watershed transform method and other watershed transformation method. Our proposed method will be more accurate. In image processing, object boundary segmentation is an important and fundamental issue.This study modified the traditional style of the watershed transformation,and we proposed the concept of an election-style, so that contact between the object image can be properly divided. By this method, image of objects can be converted according to the results of Euclidean distance transformation. It is easier to obtain images of various objects in the correct profile for easy application in medical image. Finally,we compare our methed with the National Institutes of Health (NIH) developed image processing software ImageJ, and marker-controlled watershed transformation of other scholars.

第一章 序論 1
第二章 距離轉換 4
2-1 距離轉換之定義 4
2-2 斜切式距離轉換 10
2-3 查表法與斜切式距離轉換比較 13
第三章 邊界分割的理論 17
3-1傳統的分水嶺轉換 17
3-1-1 分水嶺轉換之背景 17
3-1-2 分水嶺演算法的定義 18
3-1-3 分水嶺演算法流程描述 19
3-1-4 傳統式分水嶺之缺點 23
3-2 控制標記的分水嶺轉換 24
3-2-1 Matlab所使用的控制標記分水嶺轉換 24
3-2-2 改良後的控制標記分水嶺轉換 25
3-2-3 控制標記分水嶺之缺點 30
第四章：選舉式分水嶺轉換 31
4-1 選舉法的概念 31
4-2 先轉換後選舉 32
4-3 先選舉後轉換 35
4-4 選舉式分水嶺改善之效果 40
第五章：各種方法之間的比較 41
5-1 傳統式分水嶺轉換之實驗結果 41
5-2 ImageJ軟體說明與實驗結果 43
5-2-1 ImageJ的介紹說明 43
5-2-2 ImageJ之實驗結果 44
5-3 Matlab內控制標記分水嶺之實驗結果 47
5-4 改良式控制標記分水嶺之實驗結果 48
5-5 選舉式分水嶺之實驗結果 52
5-6 各種方式之實驗結果比較 55
第六章 結論 59
參考文獻 61

[1] Tomio Hirata , ”Aunified linear-time algorithm for computing distance maps, “ Information Processing Letters58, 129-133,1996

[2] DW.Paglieroni, GE.Ford and M.Tsujimoto, “The position-orientation masking approach to parametric search for template matching,” IEEE Trans.PAMI, 1990

[3] G.Borgefors, “Distance transformations in arbitrary dimensions,” Comput.Vision, Graphics Image Process.27, 321-345, 1984

[4] P.E.Danielsson, “Euclidean distance mapping,” Comput.Graphics Image Process.14, 227-248, 1980

[5] Heinz Breu,Joseph Gil，David Kirkpatrick, “Linear Time Euclidean Transform AlgorithmsIEEE,” Trans parrern and Machine Intelligence，Vol 17 No. 5,May 1995

[6] A.Rosenfeld and J.Pfaltz, “Sequential operation in digital picture processing,” J.Assoc.Comput. Mach. 13, 471-494, 1966

[7] DW.Paglieroni, “Distance transforms: properties and machine vision applications,” CVGIP, 1991

[8] G.Borgefors, “An improved version of the chamfer matching algorithm,” in Proc.7th Int.Conf.on Pattern Recognit., Montreal, Canada, 1175-1177, 1984

[9] G.Borgefors, “Hierachical chamfer matching: a parametric edge matching algorithm,” IEEE Trans. PAMI 10(6), 849-865, 1988
[10] DW.Paglieroni, “A unified distance transform algorithm and architecture,” Machine Vision and Applications, 47-55,1992

[11] Ingemar Ragnemalm, “Neighborhoods for Distance Transformations Using Orderd Propagation,” CVGIP: IMAGE UNDERSTANDING,Vol.56,No.3, ,399-409

[12] Ellis Horowitz and Sartaj Sahni , “Fundamentals of Data Structures in Pascal,” 3rd Ed. Computer Science Press, New York.

[13] P.E.Danielsson, “A new shape factor,” Comput.Graphics Image Proc.,292-299, 1978

[14] Weiguang Guan and Songde Ma, “A List-Processing Approach to Compute Voronoi Diagrams and the Euclidean Distance Transform,” IEEE Trans parrern and Machine Intelligence，Vol 20 No. 7,Jul. 1998

[15] L. Vincent and P. Soille, “Watersheds in digital spaces: An efficient algorithm based on immersion simulations,” IEEE Trans. Pattern Anal. Machine Intell., Vol. 13, No. 6, 583–598, Jun. 1991.

[16] N. Malpica, C. Ortiz de Solórzano, J. J. Vaquero, A. Santos, I.
Vallcorba, J. M. García-Sagredo, and F. del Pozo, “Applying watershed algorithms to the segmentation of clustered nuclei,” Cytometry,No. 28, 289–297, 1997.

[17] Xiaodong Yang, Houqiang Li, and Xiaobo Zhou,“Nuclei Segmentation Using Marker-Controlled Watershed, Tracking Using Mean-Shift, and Kalman Filter in Time-Lapse Microscopy
” IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS—I: REGULAR PAPERS, VOL. 53, NO. 11, NOV. 2006

[18] A. Bleau and L. J. Leon, “Watershed-based segmentation and region merging,” Comput. Vision Image Understanding, Vol. 77, No. 3, 317–370, 2000.

[19] R. C. Gonzalez and R. E. Woods, Digital Image Processing. Upper Saddle River, NJ: Prentice-Hall, 2002.

[20] P. S. U. Adiga and B. B. Chaudhuri, “An efficient method based on watershed and rule-based merging for segmentation of 3-D histo-pathological images,” Pattern Recogn., Vol. 34, 1449–1458, Jul. 2001.