臺灣博碩士論文加值系統

在本篇論文中，我們提出了一個快速的、以小波轉換為基礎的色彩切割技術(以下稱FWCS演算法)以及一個具方向性的區域生長技術(DRG)來處理影像中具有意義的切割。快速小波式色彩切割技術(FWCS)是在做完漸進式色彩取決的分析和以直方圖方式為基準作色彩擷取程序後再合併切割彩色區塊或碎塊。再經由計算分割出來的圖形區塊中心當成擴張種子，快速具方向性的區域生長技術(FRDG)將在一對相鄰的彩色區塊上，根據它們中心點連接的路徑方向進行方向性的區塊擴張。當兩個區塊已經經由FDRG測試並確定其連接關係之後，我們提出的系統架構會接著並計算從這兩個區塊裡取得的紋理特徵，由紋理相似度測試(TST)來更進一步地確認它們之間的關係。若任何成對的區塊能通過FDRG 與TST的雙重測試，我們就將它們視為一組具有相關聯的區域，否則它們仍然保持個別獨立。藉著使用遞迴的程序處理，當兩組關聯區塊彼此碰觸時，將繼續被融合成一個統一的區域，或者經由某方面的判定為不相連結，即判定為兩個具有高度關係的鏈結區域。我們的模擬將證明，藉由所提出的影像切割的系統架構，能夠以人類的認知感官角度來求得影像具有意義的物件或其中主要的部分。

In this thesis, we propose a fast wavelet-based color segmentation (FWCS) technique and a directional region-growing (DRG) technique towards the semantic image segmentation. The FWCS technique is a subsequent combination of progressive color analysis-by-truncation and histogram-based color extraction processes to segment color regions or fragments out. By exploring the centroids of segmented fragments as the growing seeds, FDRG runs the directional region growing on pairs of adjacent color regions according to the direction of connecting paths between those regions’ centroids. When the connection relation of two examined regions is positively confirmed by FDRG, the proposed framework subsequently compute the texture features extracted from these two regions for further checking their relation by the texture similarity testing (TST). If any pair of regions passes the double check with both FDRG and TST, we identify them as associated regions. Otherwise they are kept independent. With a recursive routine, two associated regions would be continuously either bound into a unified area while they are contacted, or recognized as two linked regions while they are somewhat separated. Our simulations demonstrate that the semantic objects or their major parts in the sense of human perception could be explored from an image through the proposed segmentation framework.

中文摘要.....................................................Ⅰ
英文摘要.....................................................Ⅱ
誌謝.........................................................Ⅲ
目錄........................................................ Ⅳ
圖目錄.......................................................Ⅵ
第一章 緒言..................................................1
1.1 研究背景...........................................1
1.2 相關技術回顧.......................................3
1.3 研究動機...........................................4
1.3 研究架構...........................................6
1.4 論文架構...........................................8
第二章 快速小波式色彩切割技術................................9
2.1 FWCS層數的決定方式................................11
2.2 FWCS低頻係數個數決定規則..........................11
2.3 低頻代表性係數之萃取..............................14
2.4 色盤對應方式......................................16
第三章 具方向性快速區域聚積演算法與紋理相似度測試...........17
3.1 具方向性快速區域聚積演算法........................17
3.2 紋理相似度測試....................................15
3.3 紋理特徵比對方式..................................19
第四章 輪廓追蹤.............................................22
4.1 輪廓追蹤..........................................22
第五章 區域間關係之策略.....................................27
5.1 區域間關係之處理程序..............................27
5.2 區域關係處理流程圖................................30
第六章 實驗結果.............................................31
6.1 FWCS與WBMT比較結果................................31
6.2 邊緣追蹤輪廓線與Sobel細線化後之結果比較...........32
6.3 鏈結方式之比較....................................34
第七章 結論.................................................37
參考文獻.....................................................39

1. S. Belongie, C. Carson, H. Greenspan, J. Malik, “Color- and texture-based image segmentation using EM and its application to content-based image retrieval, “ the 6th International Conference on Computer Vision, Page(s): 675–682, 1998.
2. C.S. Won and H. Derin, “Unsupervised segmentation of noisy and textured image using Markov random field, “ CVGIP92, Graphical models Image Processing, Vol. 54, No. 4, Page(s): 308-328, 1992.
3. S. Lakshmanan and H. Derin, “Simultaneous Parameter Estimation and Segmentation of Gibbs Random Fields Using Simulated Annealing”, IEEE Transactions on Pattern Analysis and Machine Intelligence. Vol. 11. no.8, 799-813.
4. J.F. Haddon and J.F Boyce, “Image segmentation by unifying region and boundary information, ” IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol.12, Issue: 10, Page(s): 929 –948, Oct. 1990.
5. T. Pavlidis and Y.-T. Liow, "Integrating region growing and edge detection," IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol.12, Issue: 3, Page(s): 225 –233, March 1990.
6. A. Tremeau and N. Borel, "A Region Growing and Merging Algorithm to Color Segmentation," Pattern Recognition, Vol.30, Issue: 7, Page(s): 1191-1203, July 1997.
7. A. Tremeau and P. Colantoni, ”Regions Adjacency Graph Applied to Color Image Segmentation,” IEEE Transactions on Image Processing, Vol.9, Issue: 4, Page(s): 735 -744 April 2000.
8. J. Fan, D.K. Y Yau, A. K. Elmagarmid, and W. G. Aref, “Automatic image segmentation by integration color-edge extraction and seeded region growing, ” IEEE Transactions on Image Processing, Vol.10, Issue: 10, Page(s): 1454 -1466, Oct. 2001.
9. S.A. Hojjatoleslami and J. Kittler, "Region growing: A new approach," IEEE Transactions on Image Processing, Vol.7, Issue: 7, Page(s): 1079 –1084, July 1998.
10. F. Meyer and S. Beucher, “Morphological segmentation, ”J. Vis. Commun. Image Represent, Vol.1, Page(s): 21-46, 1990.
11. Y. L. Chang; X. Li, “Adaptive image region-growing,” IEEE Transactions on Image Processing, Vol.3, Issue: 6, Page(s): 868 –872, Nov. 1994.
12. H. C. Lin, L. L. Wang, and S. N. Yang, “ Color image retrieval based on hidden Markov models, “ IEEE Transactions on Image Processing, Vol.6, No.2, Page(s): 332-339, 1997.
13. A. Gopto and R. Jain, “ Visual Information Retrieval, “ Communication of the ACM 40(5), Page(s): 69-79, 1997.
14. A. Pentland, R. W. Picard and S. S, Photobook, “ Content-Based Manipulation of Image Databases, Storage and Retrieval for Image and Video Databases II, “ San Jose, 1995.
15. J. R. Smith and S. F. Chang, “ Visually Searching the Web For Content,“ IEEE Transactions on Multimedia, Vol.3, Page(s): 12-20, 1997.
16. H. S. Malvar, “ Fast progressive wavelet coding, “ Data Compression Conference Proceedings, Page(s): 336 –343, 1999.
17. Amet, A.L. and A. Ercil, “texture Defect Detection using domain co-occurrence matrices”, Image Analysis and Interpretation, 1998 IEEE Southwest Symposium, pp.205-210,1998.
18. Haralick, R.M., Shanmugam, K. and I. Dinstein, “Textural features for image classification,” IEEE Trans. Systems Man and Cybernetics, Vol.3, pp.610-621,1973.
19. T. Chang and C.-C. J. Kuo “ Texture Analysis and Classification with Tree-Structured Wavelet Transform,” IEEE Transactions on Image Processing, Vol. 2, No. 4, pp.429-441, 1993.
20. 許嘉豪,”複雜紋理之瑕疵檢測”,國立台灣科技大學工程研究所自動化及控制學程碩士學位論文,2000.