臺灣博碩士論文加值系統

由於網路頻寬和電腦硬體快速的發展，因此大量的數位影像進入我們的生活中，像是教學(Education)、購物(Shopping)、交通管理(Traffic control)、數位圖書館(Digital library)和醫療(Medical treatment)等等，在一個資料量龐大的影像資料庫中，如何精確的取得相關的影像，便成了一個重要的議題。
通常人類對顏色的敏感度比其它特徵高，因此有許多的研究使用顏色特徵，最著名的就是顏色直方圖，其優點是能快速進行比對，但只能表示影像顏色全域性的分配，而忽略所有顏色空間關係的資訊；影像中物件的邊緣是重要的特徵，本論文提出能記錄影像中物件邊緣的顏色組合(color-pattern)的特徵，此種方式比顏色直方圖有更高的精確度。除此之外我們還抽取空間關係的特徵值(spatial-pattern)，空間特徵是抽取出影像中不同區域的圖素顏色變化，空間特徵使得搜尋的機制更豐富。
另外影像系統的回應時間(Response times)也是一個重要的指標，所以本研究提出一個圖形比對時的條件判斷機制，能提前排除不必要的影像比對，讓影像比對的速度加快，讓搜尋更有效率。

Due to the fast development of the Internet and computer hardware systems, a vast amount of digital pictures are produced. For example, Education, Shopping, Traffic control, Digit library, Medical treatment and so on. In a large image database, how to accurately retrieve images has become an important issue.
In general, humans are more sensitive to color than to shape, texture, and spatial relationship. So many researches used the color characteristic. The most popular method using the color characteristic is color histogram. The advantages of the color histogram include simple procedure and quick computation. However, color histogram can only be used to represent the global characteristics rather than the local ones in an image. The edge of the object is the important feature for the image. In this thesis, the important characteristics of two kinds of images are captured. For the color-pattern feature, neighboring color variations are recorded for the key features. It can be more accurate than color histogram. Moreover, we utilize the spatial features (spatial-pattern) to distinguish the local color variation, and this can make the search mechanism more plenty.
In addition, the response time of the image system is an important indicator. Thus, a new concept of conditional judgment is proposed. Unqualified image will be filtered without comparing the whole image. The system can search image in a more efficient way.

目錄
中文摘要 I
Abstract II
誌謝 III
目錄 IV
圖目錄 VII
表目錄 VIII
第一章、緒論 1
1.1 研究動機 2
1.2 研究目的 3
1.3 研究的範圍和限制 4
1.4 文章架構 5
第二章、文獻探討 6
2.1 顏色空間 6
2.1.1 RGB顏色空間 6
2.1.2 HSV顏色空間 7
2.1.3 CIELAB顏色空間 8
2.2 顏色特徵 9
2.2.1 傳統顏色直方圖 10
2.2.2 模糊顏色直方圖 10
2.2.3 單色與雙色的顏色直方圖 13
2.3 顏色空間特徵 17
2.3.1 顏色分群和空間分群特徵 18
2.3.2 空間色彩直方圖 20
2.3.3 顏色複雜度與顏色空間特徵 22
第三章、以顏色樣式與空間樣式為主的影像搜尋系統 25
3.1 研究架構 25
3.2特徵的取得 27
3.2.1 color-pattern特徵 27
3.2.2 spatial-pattern 特徵 30
3.3 相似度測量 30
3.3.1 color-pattern 特徵相似度測量 31
3.3.2 spatial-pattern 特徵相似度測量 34
3.4 有效率搜尋方式 34
第四章、實驗 37
4.1 color-pattern與spatial-pattern效果評估 37
4.1.1 實驗環境 37
4.1.2實驗結果的評估方式 38
4.1.3 實驗結果 39
4.2 有效率搜尋方式之實證 44
4.2.1實驗環境 44
4.2.2實驗結果 44
第五章、結論與未來工作 46
5.1 結論 46
5.2未來工作 47
參考文獻 48

圖目錄
圖1.1：颱風圖(a)：颱風距離台灣遠，(b)：颱風準備登入台灣。 3
圖1.2：影像搜尋系統架構圖。 4
圖2.1：RGB顏色立方圖。 7
圖2.2：HSV 顏色空間圓柱圖。 8
圖2.3：單色與雙色的顏色直方圖的架構圖。 14
圖2.4 梯度影像(a)影像I和三個 區塊(b)梯度影像。 15
圖2.5：HUCUB與HBCNB圖例(a)：一致區塊，(b)：不一致區塊，(c)：b圖的B和D圖素。 16
圖2.6：顏色分群與空間分群的圖例。 19
圖3.1：以顏色樣式與空間特徵為主的影像搜尋系統架構圖。 26
圖3.2：color-pattern圖例。 29
圖3.3：spatial-pattern 圖例。 30
圖4.1：實驗影像圖例。 38
圖4.2：做完303次查詢，每個方法的精確度與召回度之間的關係。 42

表目錄
表3.1：當N=9時，在不同門檻值下，最少要比較的區塊次數。 35
表3.2：當N=9和T=0.9，比較到不同區塊時，Dt的值最少要多少。 36
表4.1：每個角度取k個color-pattern特徵的平均取得精確度。 40
表4.2：每區域的每個角度取k個 color-pattern特徵的平均取得精確度。 41
表4.3：做完303次查詢，平均取得精確度。 42
表4.4：spatial-pattern對其它方法的改善率。 43
表4.5：不同門檻值下，比較的區域數。 45

參考文獻
[1]C. Carson, S. Belongie, H. Greenspan and J. Malik (2002), “Blobworld: image segmentation using expectation-maximization and its application to image querying,” IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 24, No. 8, pp. 1026-1038.

[2]Y. K. Chan and C. C. Chang (2001), “Spatial similarity retrieval in video databases,” Journal of Visual Communication and Image Representation, Vol. 12, No. 2, pp. 107-122.

[3]Y. K. Chan and C. Y. Chen (2004), “Image retrieval system based on color-complexity and color-spatial features,” The Journal of Systems and Software, Vol. 71, pp. 65-71.

[4]S. K. Chang, Q. Y. Shi and C. W. Yan (1987), “Iconic indexing by 2-D strings,” IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 9, No. 3, pp. 413-428.

[5]L. Cinque, G.. Ciocca, S. Levialdi, A. Pellicanò and R. Schettini (2001), “Color-based image retrieval using spatial-chromatic histograms,” Image and Vision Computing, Vol. 19, pp. 979-986.

[6]C. Connolly and T. Fliess (1997), “A study of efficiency and accuracy in the transformation from RGB to CIELAB color space,” IEEE Transactions on Image Processing, Vol. 6, No. 7.

[7]K. Eukunaga (1990), “Introduction to statistical pattern recognition,” Academic Press, San Diego, CA.

[8]J. M. Fuertes, M. Lucena, N. Pérez de la Blanca and J. Chamorro-Martínez (2001), “A Scheme of Colour Image Retrieval from Databases,” Pattern Recognition, Vol. 22, pp. 323-337.

[9]G. Giocca, I. Gagaliardi and R. Schettini (2001), “Quicklook2: An integrated multimedia system,” Journal of visual languages and computing, Vol. 12, pp. 81-103.

[10]R. C. Gonzalez and R. E. Woods (1992), “Digital Image Processing,” Addison-Weslesy, Reading, MA.

[11]J. Han and K. K. Ma (2002), “Fuzzy color histogram and its use in color image retrieval,” IEEE Transactions on Image Processing, Vol. 11, pp. 944-952.

[12]P. W. Huang and Y. R. Jean (1994), “Using 2D C+-Strings as spatial knowledge representation for image database systems,” Pattern Recognition, Vol. 27, No. 9, pp. 1249-1257.

[13]P. W. Huang and C. H. Lee (2004) “Image database design based on 9D-SPA representation for spatial relations,” IEEE Transactions on Knowledge and Data Engineering, Vol. 16 , pp. 1486-1496.

[14]M. S. Kankanhalli, B. M. Mehtre and H. Y. Huang (1999), “Color and spatial feature for content-based image retrieval,” Pattern Recognition Letters, Vol. 20, pp. 109-118.

[15]A. A. Khotanzad and O. J. Hernandez (2003), “Color image retrieval using multispectral random field texture model and color content features,” Patten Recognition, Vol. 36, pp. 1679-1694.

[16]W. J. Kou (2001), “Study on image retrieval and ultrasonic diagnosis of breast tumors dissertation,” Department of Computer Science and Information Engineering. National Chung Cheng University, Chiayi, Taiwan ROC.

[17]S. Y. Lee and F. J. Hsu (1992), “Spatial reasoning and similarity retrieval of images using 2D C-String knowledge representation,” Pattern Recognition, Vol. 25, No. 3, pp. 305-318.

[18]H. Nezamabadi-pour and E. Kabir (2004), "Image retrieval using histograms of uni-color and bi-color blocks and directional changes in intensity gradient," Pattern Recognition Letters, Vol. 25, pp. 1547-1557.

[19]A. K. Majumdar, I. Bhattacharya and A. K. Saha (2002), “An object-oriented fuzzy data model for similarity detection in image databases,” IEEE Transactions on Knowledge and Data Engineering, Vol. 14, No. 5, pp. 1186-1189.

[20]B. M. Mehtre, M. S. Kankanhalli and W. F. Lee (1998), “Content-based image retrieval using a composite color-shape approach,” Information Processing & Management, Vol. 34, No. 1, pp. 109-120.

[21]B. M. Mehtre, M. S. Kankanhalli, A. D. Narasimhalu and G. C. Man (1995), “Color matching for image retrieval,” Pattern Recognition Letter, Vol. 16, pp. 325-331.

[22]G.. Qiu (2002), “Indexing chromatic and achromatic patterns for content-based colour image retrieval,” Pattern Recognition, Vol. 35, pp. 1675-1686.

[23]G. Paschos (2001), “Perceptually uniform color spaces for color texture analysis: An empirical evaluation,” IEEE Transactions on Image Processing, Vol. 10, No. 6, pp. 932-937.

[24]H. Permuter, J. Francos and I. Jermyn (2006), “Astudy of Gaussian mixture models of color and texture features for image classification and segmentation,” Pattern Recognition, Vol. 39, pp. 695-706.

[25]E. Petrakis, C. Faloutsos and K. I. Lin (2002), “ImageMap: An image indexing method based on spatial similarity,” IEEE Transactions on Knowledge and Data Engineering, Vol. 14, No. 5, pp. 979-987.

[26]A. Rao, R. K. Srihari, L. Zhu and A. Zhang (2002), “A method for measuring the complexity of image databases,” IEEE Transactions on Multimedia, Vol. 4, No. 2, pp. 160-173.

[27]E. Şaykol, U. Güdükbay and Ö. Ulusoy (2005), “A histogram-based approach for object-based query-by-shape-and-color in image and video databases,” Image and Vision Computing, Vol. 23, pp. 1170-1180.

[28]A. F. Sintans, J. M. Cadenas and F. Martin (1999), “Membership functions in the fuzzy C-Means algorithm,” Fuzzy Sets and Systems, Vol. 101, pp. 49-58.
[29]M. C. Su and C. H. Chou (2001), “A modified version of the k-means algorithm with a distance based on cluster system,” IEEE Transactions on Pattern Analysis and Machine Intelligence , Vol. 23, No. 6, pp. 674-680.

[30]J. Sun, X. Zhang, J. Cui and L. Zhou (2006), “Image retrieval based on color distribution entropy,” Pattern Recognition Letters, Vol. 27, pp. 1122-1126.

[31]M. J. Swain and D. H. Ballard (1991), “Color indexing,” International Journal of Computer Vision, Vol. 7, No. 1, pp. 11-32.

[32]Y. H. Wang (2003), “Image indexing and similarity retrieval based on spatial relationship model,” Information Science, Vol. 154, pp. 39-58.

[33]C. H. Yan and S. Y. Chen (2002), “Retrieval of translated, rotated and scaled color textures,” Pattern Recognition, Vol. 36, pp. 913-929.

[34]H. W. Yoo, H. S. Park and D. S. Jang (2005), “Expert system for color image retrieval,” Expert System with Applications, Vol. 28, pp.347-357.

[35]H. Zhai, P. Chavel, Y Wang, S. Zhang and Y. Liang (2005), “Weighted fuzzy correlation for similarity measure of color-histograms,” Optics Communications, Vol. 247, pp. 49-55.