在數位時代中，影片資料在生活中也越來越普及。當使用者與影片資料量越來越多時，對於影片資料之管理也越來越重要。因此影片資料庫系統之實現，提供使用者查詢並擷取影片資料。
本論文使用碎形正交基底編碼(Fractal orthonormal bases)技術結合支援向量分類(Support vector clustering)找出場景變化處，在從各場景中找出各別的關鍵影格作為資料庫索引，建立影片資料庫，每張資料庫內影像之特徵均由對碎形正交基底之投影向量值表示。正交基底是由碎形迭代函數透過target及domain blocks比對所訓練導出，可證明相似影像具相似碎形函數，而且不相似影像具相異碎形特徵向量；換言之，特徵點相距越遠，保證其對應影像內容一定不相似，然而特徵點較靠近，則保證其影像內容相似。因此，使用碎形正交基底函數線性組合所得係數為搜尋資料庫索引鍵值，可取得相似影像，並避免找出不相似影像。
由於欲搜尋之影像很難根據單一張搜尋影像(query image)代表所有可能之形狀、大小或方位，為使搜尋條件更為明確，藉由輸入多張與目標影像正、負相關搜尋影像，透過Multiple-Instance learning 法則自動地找出與正相關影像(positive examples) 相似且與負相關(negative examples)不相似之碎形正交基底投影向量特徵，使搜尋條件更為明確，將使用者最有興趣之部分，結合具有良好索引檔之碎形正交基底之技術。
影像比對時，方法是依據MIL所擷取之特徵，找尋資料庫哪些影像具有相似特徵，計算相似度，依此作排名輸出。詳細比對時，將資料庫中有著搜尋特徵之影像，找出該所屬區域，將擷取之特徵群正規化，求得每個特徵群佔所有搜尋特徵群之比例關係，再以依正相關特徵群之比例和資料庫影像特徵群比例，類似計算histogram之方式求得特徵比例相似度之外；另外還加入計算所求得特徵群之間結構關係，與正相關範例影像之特徵群結構關係亦計算特徵結構相似度；在加入每個特徵群區域之分散程度，及簡單計算其區域變異數亦和正相關範例做比較，於上述三者加入相似性量測中。

During the Digital Period, the more people using these digital video. When there are more and more users and amount of video data, the management of video data becomes a significant dimension during development. Therefore, there are more and more studying of accomplishing video database system, which provide users to search and get them.
In this paper, a novel method for Video Scene Change Detection and video database retrieval is proposed. Uses Fractal orthonormal bases to guarantee the similar index has the similar image the characteristic union support vector clustering, splits a video into a sequence of shots, extracts a few representative frames(key-frames) to take the video database index from each shot.
When image search compared to, according to MIL to pick up the characteristic, which images pursues the video database to have the similar characteristic, computation similar, makes the place output according to this.

摘要 i
目錄 iii
圖目錄 vi
表目錄 x
第1章 簡介 1
1.1 相關研究(shot detection) 9
1.1.1 以相鄰影格差值進行分析 9
1.1.2 以色彩直方圖 (color histogram)進行分析 12
1.1.3 以邊緣像素(edge pixel)進行分析 14
1.1.4 以相似度比對(Likelihood ratio)進行分析 16
1.1.5 以線性回歸方法偵測 16
1.2 相關研究(retrieval) 18
1.2.1 以顏色為基礎之擷取概念 19
1.2.2 以形狀為基礎之擷取概念 22
1.2.3 以內容為基礎之擷取概念 25
1.3 影片搜尋之相關研究 27
1.3.1 JUST A CONTENT-BASED QUERY SYSTEM FOR VIDEO DATABASES 27
1.3.2 Fast Image/Video Retrieval On Compressed Image And Video Databases 30
第2章 理論基礎 34
2.1 碎形理論 34
2.1.1 轉換之收歛性 37
2.1.2 迭代函數系統 (iterative function system) 37
2.1.3 影像分割 38
2.1.4 迭代函數 39
2.1.5 碎形在影像搜尋的應用 42
2.1.6 Orthogonal Basis IFS 44
2.2 支援向量分類 ( Support vector clustering ) 49
2.3 Multiple-Instance Learning 57
2.3.1 定義 58
2.3.2 MIL應用於影像搜尋 58
2.3.3 Diverse Density Algorithm 61
2.3.4 Diverse Density definition 62
2.3.5 計算 63
2.3.6 計算 65
2.3.7 Finding the maximum 65
第3章 研究方法步驟及結果 67
3.1 特徵分析及資料庫建立 69
3.1.1 空間轉換 69
3.1.2 資料分類 72
3.1.3 資料庫建立 80
3.2 影像搜尋 83
3.2.1 使用MIL找出共有特徵 83
3.2.2 比對方法 85
3.3 實驗結果 88
參考文獻 92

[1] Fernando W.A.C., Canagarajah C.N., Bull D.R., “Fade and dissolve detection in uncompressed and compressed video sequences,” in Proc. of the IEEE. ICIP, pages 299–303, 1999.
[2] Ba Tu Truong, Chitra Dorai, Svetha Venkatesh, “New enhancements to cut, fade, and dissolve detection processes in video segmentation”, in Proc. ACM International Conf. Multimedia, p. 219–227, 2000.
[3] H. J.W. Zhang, A. Kankanhalli, and S. Smoliar, “Automatic partitioning of full-motion video,” Multimedia Syst., vol. 1, no. 1, pp. 10–28, 1993.
[4] Ramin Zabih, Justin Miller, Kevin Mai, “A feature-based algorithm for detecting and classifying production effects,” in Proc. ACM Multimedia San Francisco, CA, 1995, pp. 189-200.
[5] R. Kasturi, R. Jain, ”Dynamic Vision,” Computer Vision: Principles, Eds. R. Kasturi, R. Jain, IEEE Computer Society Press, Washington, 1991, pp. 469-480.
[6] Seung-Hoon Han, In-So Kweon, “Detecting cuts and dissolves through linear regression analysis,” Electronics Letters, Volume 39, Issue 22, 30 Oct. 2003 Page(s):1579 – 1581
[7] Smith, J.R., and Chang, S.F., “VisualSEEk: a fully automated content-based image query system,” in Proceedings ACM Multimedia, Boston MA, ACM Press, 87-98, 1996
[8] Chad Carson, “Blobworld: a system for region-based image indexing and retrieval,” International conference on visual information system, 1999.
[9] D. Androutsos, K.N. Plataniotis, and A.N. Ventsanopoulos, “A novel Vector-based approach to color image retrieval using a vector angular-based distanced measure,” Computer Vision and Image Understanding, vol. 75, no. 1/2, pp. 46-58, 1999.
[10] Yu, H.H, "Fast image/video retrieval on compressed image and video databases" Multimedia Signal Processing, 1999 IEEE 3rd Workshop on 13-15 Sept. 1999 Page(s):129-134
[11] La Cascia, M.; Ardizzone, E. "JACOB: just a content-based query system for video databases" Acoustics, Speech, and Signal Processing, 1996. ICASSP-96. Conference Proceedings., 1996 IEEE International Conference on Volume 2, 7-10 May 1996 Page(s):1216 - 1219 vol. 2
[12] B. Mandelbrot, The Fractal Geometry of Nature, San Francisco, CA: Freeman,1982.
[13] M. F. Barnsley, Fractals Everywhere, Academic Press, San Diego, 1988.
[14] Arnaud E. Jacquin, “A Fractal Theory of Iterated Markov Operators with Applications to Digital Image Coding” , Ph.D. Thesis, Georgia Institute of Technology, 1989.
[15] Yuval Fisher, E. W. Jacobs, and R.D. Boss, “Iterated Transformation image compression,” NOSC Thec. Rep. TR-1408, Naval Oceans Systems Center, San Diego, CA, 1991.
[16] Yuval Fisher, “Fractal image compression : theory and application” , Springer, New York, 1996.
[17] A.E. JACQUIN, “Image Coding based on a fractal theory of iterated contractive image transformation,” IEEE Trans. on Image Process. ,1992.
[18] Z.Wang, Z.Chi and D.Feng, “Content-based image retrieval using block-constrained fractal coding and nona-tree decomposition,” IEEE Signal Processing,2000.
[19] G. Vines and M. H. Hayes, “Nonlinear interpolation in a one-dimensional fractal model,” In Proceedings of the fifth Digital Signal Processing Workshop, pp. 8.7.7-8.7.2, 1992.
[20] G. Vines and M.H. Hayes, “Nonlinear Address Maps in a One-Dimensional Fractal Model,” IEEE Trans. on Signal Processing, 1993.
[21] G. Vine, “Signal Modeling with Iterated Function System,” PhD thesis, Georgia Institute of Technology, Atlanat, GA, 1993.
[22] V. Vapnik, “Estimation of dependencies based on empirical data,” Springer-Verlag, New York, 1982
[23] Asa Ben-Hur, David Horn, Hava T. Siegelmann, Vladimir Vapnik, “A support vector clustering method,” International Conference on Pattern Recognition (ICPR'00)-Volume 2, September 03-08, 2000.
[24] D. Tax, R. Duin., “Support vector domain description,” Pattern Recognition letters, 20:1991-1999, 1999.
[25] O. Maron and A. Lakshmi Ratan, “Multiple-instance learning for natural scene classification,” in Machine Learning: Proc. 15th International Conference, 1998.