在影像辨識的領域及數位版權管理中的浮水印，局部不變性特徵點已是被廣泛使用的方法，而尺寸不變性特徵轉換法是最常被使用來截取特徵點，此方法所取出的特徵點用於不同影像之間的辨識具有相當的可靠性，並且，這些特徵點能抵抗影像轉檔、尺寸縮放、旋轉等變化；然而，大量的特徵點對辨識階段將是一個重要的問題，影像的匹配會因為影像的數量而增加。
我們提出一個方法，利用二元化精簡的表示尺寸不變性特徵。我們也發現二元化的尺寸不變性特徵可以維持很高的匹配準確率；此外，可以用精簡易於擴充的雜湊方法有效率的建立二元化尺寸不變性特徵的索引(例如dbm及其他dbm形態)。我們在較大的資料庫裡(大約1400萬個特徵點)，實作一個尺寸不變性特徵的匹配引擎來證實我們的假設，結果指出我們提出的方法在影像批配的應用中，可以有效率的減輕因為資料量的擴充所產生的問題。

The local invariant features have been widely used in the image recognition and in the Digital Right Management (DRM) applications. Among all the local invariant features, the Scale Invariant Feature Transform (SIFT) is one of the most adopted method. The feature keypoints extracted by the SIFT are invariant to image translation, scaling, rotation, etc., and hence a promising technique to identify plagiarism among images. However, the large number of local invariant feature keypoints extract by the SITF poses a scalability problem in the matching stage as the number of images to be matched increases.
We propose a method to simplified the SIFT descriptor through binarization. We discover that the binarized SIFT descriptors still maintain very high distinctiveness for the image matching process. Moreover, the binarized SIFT descriptors can be indexed by simple extensible hashing method (e.g., Dbm and its variations) very efficiently. We verify our hypothesis by implementing a SIFT descriptors matching engine with a large scale of data (> 14M keypoints). The results suggest that the proposed method can effectively alleviate the scalability problem in image matching applications.

[1] D. G. Lowe. Object recognition from local scale-invariant features. Proceedings of International Conference on Computer Vision, pages. 1150–1157, 1999.
[2] D. G. Lowe. Distinctive image features from scale-invariant keypoints. International Journal of Computer Vision, pages.91-110, 2004.
[3] Y. Ke and R. Sukthankar. PCA-SIFT: A More Distinctive Representation for Local Image Descriptors. IEEE Computer Society Conference on Computer Vision and Pattern Recognition, pages. 511-517, 2004.
[4] Y. Rui, T.S. Huang, and S. Chang, “Image Retrieval: Current Techniques, Promising Directions and Open Issues,” Visual Comm. and Image Representation, vol. 10, no. 4, pages. 39-62, 1999.
[5] R. C. Veltkamp and M. Tanase. Content-Based Image Retrieval Systems: A Survey. Technical Report UU-CS-2000-34, Dept. of Computing Science, Utrecht University, 2000.
[6] Ritendra Datta , Jia Li , James Z. Wang. Content-based image retrieval: approaches and trends of the new age. Proceedings of the 7th ACM SIGMM international workshop on Multimedia information retrieval, pages. 253-262, 2005.
[7] A. E. Abdel-Hakim and A. A. Farag. CSIFT: A SIFT descriptor with color invariant characteristics. IEEE Computer Society Conference on Computer Vision and Pattern Recognition, pages. 1978–1983, 2006.
[8] J.M. Geusebroek, R. van den Boomgaard, A.W.M. Smeulders, and H. Geerts. Color Invariance. IEEE Trans. Pattern Analysis and Machine Intelligence, vol. 23, no. 12, pages. 1338-1350, 2001.
[9] Berkeley database, http://www.oracle.com/technology/products/berkeley-db/index.html
[10] I.T. Jollife. Principal Component Analysis. New York: Springer-Verlag, 1986.
[11] V Gaede, O Gunther. Multidimensional Access Methods. ACM Computing Surveys, vol. 30, no. 2, pages. 123-169, 1998.
[12] A. Guttman. R-Trees: A Dynamic Index Structure for Spatial Searching. Proc. ACM SIGMOD Conf. Management of Data, pages. 47-57, 1984.
[13] Jon Louis Bentley. Multidimensional binary search trees used for associative searching. Communications of the ACM, vol.18 no. 9, pages.509-517, 1975.
[14] Piotr Indyk, Rajeev Motwani. Approximate nearest neighbors: towards removing the curse of dimensionality. Proceedings of the thirtieth annual ACM symposium on Theory of computing, pages. 604-613, 1998.
[15] Alex Andoni’s LSH page, http://web.mit.edu/andoni/www/LSH/index.html
[16] Yan Ke , Rahul Sukthankar. Larry Huston, An efficient parts-based near-duplicate and sub-image retrieval system. Proceedings of the 12th annual ACM international conference on Multimedia, pages. 869-876 , 2004.
[17] A. Gionis, P. Indyk, and R. Motwani. Similarity search in high dimensions via hashing. Proceedings of International Conference on Very Large Databases, pages. 518-529, 1999.
[18] JDBM, http://jdbm.sourceforge.net/
[19] H. Y. Lee, H. Kim, and H.K. Lee. Robust image watermarking using local invariant features. Optical Engineering, v.45 n.3, 037002, 2006.