Algorithms for Indexing and Integrating MPEG-7 Visual Descriptors

MPEG-7 시각 정보 기술자의 인덱싱 및 결합 알고리즘

  • Published : 2007.01.15

Abstract

This paper proposes a new indexing mechanism for MPEG-7 visual descriptors, especially Dominant Color and Contour Shape descriptors, that guarantees an efficient similarity search for the multimedia database whose visual meta-data are represented with MPEG-7. Since the similarity metric used in the Dominant Color descriptor is based on Gaussian mixture model, the descriptor itself could be transform into a color histogram in which the distribution of the color values follows the Gauss distribution. Then, the transformed Dominant Color descriptor (i.e., the color histogram) is indexed in the proposed indexing mechanism. For the indexing of Contour Shape descriptor, we have used a two-pass algorithm. That is, in the first pass, since the similarity of two shapes could be roughly measured with the global parameters such as eccentricity and circularity used in Contour shape descriptor, the dissimilar image objects could be excluded with these global parameters first. Then, the similarities between the query and remaining image objects are measured with the peak parameters of Contour Shape descriptor. This two-pass approach helps to reduce the computational resources to measure the similarity of image objects using Contour Shape descriptor. This paper also proposes two integration schemes of visual descriptors for an efficient retrieval of multimedia database. The one is to use the weight of descriptor as a yardstick to determine the number of selected similar image objects with respect to that descriptor, and the other is to use the weight as the degree of importance of the descriptor in the global similarity measurement. Experimental results show that the proposed indexing and integration schemes produce a remarkable speed-up comparing to the exact similarity search, although there are some losses in the accuracy because of the approximated computation in indexing. The proposed schemes could be used to build a multimedia database represented in MPEG-7 that guarantees an efficient retrieval.

본 논문에서는 MPEG-7 시각 정보 기술자인 Dominant Color와 Contour Shape 기술자에 대한 새로운 인덱싱 알고리즘을 제안한다. Dominant Color 기술자에서 사용되는 비교 연산 식은 가우스 혼합 모델에 기초하고 있기 때문에 기술자의 각 속성들을 하나의 칼라 히스토그램 형태로 변형시켜서 인덱스로 사용한다. Contour Shape 기술자는 두 단계 형태의 알고리즘을 사용하는데, 첫 번째 단계에서는 글로벌 변수인 Eccentricity와 Circularity를 사용한 대략적인 비교를 통해서 비슷하지 않은 이미지 오브젝트를 배제시키고 두 번째 단계에서 남겨진 오브젝트들과 질의 오브젝트들간의 Peak 변수를 사용한 비교 연산을 통해 인덱싱을 수행한다. 또한 본 논문은 효율적인 멀티미디어 데이타 검색을 위해서 두 가지의 MPEG-7 시각 정보 기술자 결합 알고리즘을 제안한다. 첫 번째 결합 알고리즘은 가중치를 확률로 변환해서 반영하는 것이고 두 번째는 가중치를 각 비교 연산 결과값의 중요도로 간주하는 방법이다. 실험을 통해서 결과를 분석해 보면 근사화를 통한 인덱스 생성으로 100%의 정확도를 유지 할 수는 없지만 논문에서 제안된 각 기술자의 인덱싱 알고리즘과 기술자들의 결합 알고리즘은 기본 검색 알고리즘과 비교했을 때 매우 빠른 속도 향상을 보여주었다. 본 논문에서 제안된 알고리즘은 MPEG-7을 사용하는 검색 시스템의 데이타베이스 구축에 효율적으로 사용될 수 있다.

Keywords

References

  1. B.S. Manjunath, Philippe Salembier and Thomas Sikora, Introduction to MPEG-7 Multimedia Content Description Interface, JOOHN WILEY & SONS, 2002
  2. ISO/IEC JTC 1/SC 29/WG 11, Information Technology - Multimedia Content Description Interface - Part 3:Visual, 2001
  3. ISO/IEC JTC1/ SC29/WG11/ N4031, Overview of the MPEG-7 Standard (version 5.0), March 2001
  4. R. Weber, H.J. Schek, and S. Blott, 'A Quantitative Analysis and Performance Study for Similarity-Search Methods in High-Dimensional Spaces,' Proceedings of ICVLDB, pp.194-205, 1998
  5. G.H. Cha, X. Zhu, D. Petkovic, and C.W. Chung, 'An Efficient Indexing Method for Nearest Neighbor Searches in High-Dimensional Image Databases,' IEEE Transaction on Multimedia, Vol. 4, No. 1, pp. 76-87, 2002 https://doi.org/10.1109/6046.985556
  6. Jinguk Jeong and Jongho Nang, 'An Efficient Bitmap Indexing Method for Similarity Search in High Dimensional Database,' ICME2004 제출중
  7. N. Beckmann, H.P, Kriegal, R. Schneider, and B. Seeger, 'The R*-tree: An Efficient and Robust Access Method for Points and Rectangles,' Proceedings of ACM SIGMOD Conference, pp. 322-331, 1990
  8. S. Berchtold, D.A. Keim, and H.P. Kriegel, 'The X-tree: An Index Structure for High-Dimensional Data,' Proceedings of 22nd International Conference Very Large Databases, pp. 28-39, 1996
  9. H. Lu, Y. Yeung Ng, and Z. Tian, 'T-tree or B-tree : Main Memory Database Index Structure,' Proceedings of 12th Australasian Database Conference, pp. 263-270, 2001
  10. N. Katayama, and S. Satoh, 'The SR-tree: An Index Structure for High-Dimensional Nearest Neighbor Queries,' Proceedings of ACM SIGMOD Conference, pp. 369-380, 1997 https://doi.org/10.1145/253262.253347
  11. C.C. Aggarwal, and P.S. Yu, 'The IGrid Index: Reversing the Dimensionality Curse for Similarity Indexing in High Dimensional Space,' Proceedings of ACM SIGKDD International Conference, pp. 119-129, 2000
  12. E. Tuncel, H. Ferhatosmanoglu, and K. Rose, 'VQ-Index: An Index Structure for Similarity Searching in Multimedia Databases,' Proceedings of ACM International Conference on Multimedia, pp. 543-552, 2002
  13. K.V.R. Kanth, D. Agrawal, and A. Singh, 'Dimensionality Reduction for Similarity Searching in Dynamic Databases,' Proceedings of ACM SIGMOD Conference, pp. 166-176, 1998 https://doi.org/10.1145/276305.276320