臺灣博碩士論文加值系統

隨著網路快速的發展與大量的數位影像資訊，越來越多的圖片應用在商業的網站上，如何有效且快速的從資料庫中搜尋這些數位資訊已經成為一個重大的議題。而現今大部分網站上的商品圖片多半是有附帶文字說明或是標籤，傳統上的作法是把圖片標上一些關鍵字，使得使用者可依據所下的關鍵字來找到商品。但是，有些使用者想要找的商品可能不容易使用關鍵字來搜尋，或是使用者不知道如何去形容商品。像是服飾商品中，常常包含著一些難以使用常見的關鍵字來描述的特徵，像是紋理(texture)、外型(shape)或是一些微小的細節。為了解決這方面的問題，我們開發出一個使用圖像為基礎的服飾檢索系統，可讓使用者使用以圖找圖的方式檢索服飾。
在本篇論文中，我們提供藉由外型及樣式的視覺化搜尋，使用了“Shape Context”做為外型的特徵描述子，並且加入文本檢索(text retrieval)的方式提昇準確率。使得使用者能夠對於相似外型的服飾進行有效的搜尋。基於外型的檢索分為兩個步驟，第一，擷取服飾上適合的特徵描述。第二，建立服飾影像的搜索策略。我們實驗中的服飾資料庫包含多種不同樣式，經由實驗，我們證明我們的檢索方法有助於服飾商品外型及樣式的檢索。

With a rapidly increasing use of the Internet, there is a large amount of digital image information; more and more picture information is applied to the commercial website. How to retrieve the digital image in database efficiently has been an significant issue. Besides described by words, pictures of clothes products are often attached. Traditionally, we tag each image with several keywords, so that users can search keywords to find products. However, the keywords may contain no clue of what they are looking for. Clothes products often contain many characteristics which are hard to describe in regular keywords, such as texture, shape, or any other subtle details. To tackle this issue, we develop the image-based visual clothing search system, which enables users to retrieve clothing images.
In this work we provide the visual search with shape using the "Shape Context" to be the shape descriptor and provide a method from text-retrieval technology to increase accuracy so that one can search for similar shapes of clothes efficiently. We develop the system with two steps. First, appropriate feature description from clothes image is extracted . Second, retrieval strategies of clothes image in database is built. And, we experiment on an image database with several kinds of clothes type, and the experiment shows that our approach is more helpful for ranking clothes product by shape and style.

1 Introduction
1.1 Issues Addressed
1.2 Contributions
1.3 Organization of the thesis
2 Related Work
2.1 Visual Vocabulary and Text-Retrieval Method
2.2 Visual Internet Object Search (VIOS) system
2.3 Garment Visual Search based on Shape Context
2.4 Summary
3 Overview of Framework
3.1 Database Images
3.2 Foreground Extraction
3.3 Feature Extraction and Indexing
3.4 Ranking Images
4 Shape Feature and Indexing
4.1 Local Feature Extraction
4.2 Shape Context
4.3 Visual Vocabulary Construction
4.4 Forward File and Inverted table
4.5 Feature Weighting and Ranking Mechanism
5 Experiment and Result
5.1 Queries
5.2 Performance and Retrieval Example
5.2.1 Evaluate different weighting methods
5.2.2 Evaluate different visual vocabulary sizes
5.2.3 Evaluate different retrieval method
5.2.4 Evaluate the influence on different soft-weighting parameters
5.2.5 Measure the accuracy of Top-3 hit rate
5.2.6 Examples of shape query and retrieval
5.3 Summary
6 VIOS system
6.1 Ranking Method for Combined Retrieval
6.2 Combined VIOS system with shape function
6.3 Comparison of VIOS system and Google Similar Images
7 Conclusion and Future Work
A K-means Algorithm

[1] S. Belongie and J. Malik, "Matching with Shape Contexts", IEEE proc.
Content-Based Access of Image and Video Libraries, 2000, pp. 20-26.
[2] A. W. M. Smeulders, M. Worring, S. Santini, A. Gupta, R. Jain,
"Content-Based Image Retrieval at the End of the Early Years", IEEE
Trans. Pattern Anal. Machine Intel., vol. 22, no. 12, 2000.
[3] R. Datta, D. Joshi, J. Li, and J. Z. Wang. "Image retrieval: Ideas, in
uences, and trends of the new age", ACM Computing Surveys, 2008.
[4] D. Lowe, "Distinctive Image Feature from Scale-Invariant Keypoints",
International Journal of Computer Vision, vol. 60, no. 2, pp. 91-110,
2004.
[5] M. Bober, "MPEG-7 Visual Shape Descriptors", IEEE Trans. on Circuits
and Systems for Video Technology, pp. 716-719, 2001.
[6] G. Mori, S. Belongie and J. Malik, "Shape Contexts Enable Efficient
Retrieval of Similar Shapes", Proc. IEEE Comp. Soc. Conf. Computer
Vision and Pattern Recognition, vol. 1, 2001, pp. 723-730.
[7] S. Belongie, J. Malik, and J. Puzicha, "Shape Matching and Object
Recognition Using Shape Contexts", IEEE Transactions on Pattern
Analysis and Machine Intelligence, 2002.
[8] C.C. Lin, C.T. Chang, "A Fast Shape Context Matching Using Indexing",
International Conference on Genetic and Evolutionary Computing, 2011.
[9] C. Rother, V. Kolmogorov, A. Blake, "GrabCut - Interactive Foreground
Extraction using Iterated Graph Cuts". ACM Transactions on Graphics
(TOG), vol. 23, pp. 309-314, 2004.
[10] J. Sivic and A. Zisserman, "Efficient Visual Search for Objects in
Videos", Proc. IEEE , Vol. 96, pp. 548-566, 2008.
[11] G. Csurka, C. Bray, C. Dance, and L. Fan, "Visual Categorization with
Bags of Keypoints", In Proc. Workshop Statistical Learning Computer
Vision, ECCV , pp. 1-22, 2004.
[12] K.C. Chang, J.J. Tsay, "A Study of Visual Clothing Search", CSIE
department, Chung Cheng University, ROC , 2010.
[13] J.J. Tsay, C.H. Lin, C.H. Tseng, K.C. Chang, "On Visual Clothing
Search", Technologies and Applications of Arti¯cial Intelligence , 2011.
[14] C.H. Tsen, S.S. Hung, J.J. Tsay, Derchian Tsaih, "An Efficient Garment
Visual Search based on Shape Context", WSEAS TRANSACTIONS ON
COMPUTERS , pp. 1195-1204, 2009.
[15] G. Mori, S. Belongie, J. Malik, "Efficient Shape Matching Using Shape
Context", IEEE, PAMI , 27(11), 2005.
[16] J. Tangelder, R. C. Veltkamp, "A survey of content based 3D shape
retrieval methods", Proceedings Shape Modeling International, pp. 145-
156, 2004.
[17] I. Elsayad, J. Martinet, T. Urruty, C. Djeraba, "A New Spatial Weight-
ing Scheme for Bag-of-Visual-Words", CBMI , pp. 1-6, 2010.
[18] J. Sivic and A. Zisserman, "Video Google: A Text Retrieval Approach
to Object Matching in Videos", in ICCV , 2003.
[19] J. Zhang, M. Marszalek, S. Lazebnik and C. Schmid, "Local Features
and Kemels for Classi¯cation of Texture and Object Categories: A Com-
prehensive Study", IJCV, vol. 73, no. 2, pp. 213-238, 2007.
[20] Y. G. Jiang, C. W. Ngo, and J. Yang, "Towards Optimal Bag-of-Features
for Object Categorization and Semantic Video Retrieval", in ACM CIVR,
2007.
[21] Google Similar Images, http://similar-images.googlelabs.com/
[22] Google Image Search, http://www.google.com/imghp
[23] Y. Boycov, M. Jolly, "Interactive graph cuts for optimal boundary and
region segmentation of objects in N-D images", In Proc. IEEE Int. Conf.
on Computer Vision, CDVROM, 2001.