臺灣博碩士論文加值系統

影像縮放技術在生活中廣泛地應用，但標準比例縮放方法往往會造成物體受到幾何變形，相對於標準比例縮放，基於內容的影像縮放方法-Seam Carving，能夠考慮影像內容，避免影像幾何變形。但對於顏色單調的物體，Seam Carving會造成物體扭曲，因此我們結合Seam Carving與標準比例縮放的特性，提出“比例縮放扭曲物體”與“先Seam Carving後比例縮放”的方法，能夠避免顏色單調的物體被扭曲，藉由我們提出的改善方式，能夠提高Seam Carving的適用性，讓影像有更好的縮放結果。

Image resizing has been widely used in life. However, standard image scaling often causes objects to be distorted. Compare to standard image scaling, another method called seam carving that supports content-aware image resizing .It can resizes images without distortion. But for the objects which color is monotonous, Seam Carving still would cause objects to be distorted. Therefore, according to the characteristics of Seam Carving and standard image scaling, we propose our ideas, “scale warped objects” and “first seam carving, last scaling”. With the improvement, it can avoid the objects which color is monotonous being distorted and Seam Carving will be more robust.

內容目錄
口試委員審定書 I
誌謝 II
摘要 III
ABSTRACT IV
圖表目錄 VII
CHAPTER 1. 緒論 1
1.1 背景 1
1.2 動機 3
1.3 目標與貢獻 4
1.4 章節介紹 4
CHAPTER 2. 相關技術 5
2.1 SEAM CARVING 5
2.1.1 原理 5
2.1.2 演算法 5
2.1.3 Seam Carving效果 10
CHAPTER 3. 問題分析與改善 13
3.1問題分析 13
3.1.1問題1-背景扭曲 13
3.1.2問題2-前景扭曲 16
3.2改善演算法 18
3.2.1問題1改善方式-比例縮放扭曲物體 18
3.2.2問題2改善方式-先Seam Carving後比例縮放 23
CHAPTER 4. 改善結果評估 26
4.1改善成果-比例縮放扭曲物體 26
4.2改善成果-先SEAM CARVING後比例縮放 27
CHAPTER 5. 結論與未來工作 28
5.1 結論 28
5.2 未來工作 29
圖片版權 30
參考資料 31

[1]."SIOX: Simple Interactive Object Extraction." from HUhttp://www.siox.org/UH.

[2].Shai, A. and S. Ariel (2007). "Seam carving for content-aware image resizing." UACM Trans. Graph.U 26(3): 10.

[3].Bertalmio, M., L. Vese, et al. (2003). "Simultaneous structure and texture image inpainting." UIEEE Trans Image ProcessU 12(8): 882-9.

[4].Criminisi, A., P. Perez, et al. (2003). UObject removal by exemplar-based inpaintingU. Computer Vision and Pattern Recognition, 2003. Proceedings. 2003 IEEE Computer Society Conference on.

[5].Friedland, G., K. Jantz, et al. (2005). USIOX: simple interactive object extraction in still imagesU. Multimedia, Seventh IEEE International Symposium on.

[6].Iddo, D., C.-O. Daniel, et al. (2003). "Fragment-based image completion." UACM Trans. Graph.U 22(3): 303-312.

[7].Viola, P. and M. Jones (2001). URapid object detection using a boosted cascade of simple featuresU. Computer Vision and Pattern Recognition, 2001. CVPR 2001. Proceedings of the 2001 IEEE Computer Society Conference on.