Video Index Generation and Search using Trie Structure

Trie 구조를 이용한 비디오 인덱스 생성 및 검색

  • 현기호 (영산대학교 컴퓨터정보공학부) ;
  • 김정엽 (영산대학교 멀티미디어공학부) ;
  • 박상현 (포항공과대학교 컴퓨터공학과)
  • Published : 2003.08.01

Abstract

Similarity matching in video database is of growing importance in many new applications such as video clustering and digital video libraries. In order to provide efficient access to relevant data in large databases, there have been many research efforts in video indexing with diverse spatial and temporal features. however, most of the previous works relied on sequential matching methods or memory-based inverted file techniques, thus making them unsuitable for a large volume of video databases. In order to resolve this problem, this paper proposes an effective and scalable indexing technique using a trie, originally proposed for string matching, as an index structure. For building an index, we convert each frame into a symbol sequence using a window order heuristic and build a disk-resident trie from a set of symbol sequences. For query processing, we perform a depth-first search on the trie and execute a temporal segmentation. To verify the superiority of our approach, we perform several experiments with real and synthetic data sets. The results reveal that our approach consistently outperforms the sequential scan method, and the performance gain is maintained even with a large volume of video databases.

비디오 데이타베이스에서 유사도 정합은 비디오 클러스터링과 비디오 라이브러리 등과 같은 많은 새로운 응용분야에서 중요성이 증가하고 있다. 대용량 데이타베이스에서 효과적인 접근을 제공하기 위하여 다양한 공간과 시간에 대한 특징치를 이용한 비디오 인덱싱 분야의 많은 연구노력이 있어왔다. 그러나 대부분의 기존 방법들은 순차적인 정합방법 또는 메모리 기반의 역 파일 기법 등에 의존하므로 대용량 데이타베이스에는 적합하지 않다. 이러한 문제를 해결하기 위하여 본 논문에서는 효과적이고 스케일 조정가능한 인덱싱 기법을 제안하기 위하여, 문자열 정합을 위해 제안된 trio를 인덱스 구조로 이용하였다. 인덱스 구성을 위하여 윈도우 순서 휴리스틱을 이용하여 각 프레임을 기호 시퀀스로 변환하고, 기호 시퀀스의 집합으로부터 디스크 상주 trio를 구성하였다 질의 처리를 위하여 trio 상에서 깊이-우선 검색과 시간 축분할을 실시하였으며, 제안한 방법의 성능을 검증하기 위하여 실제와 합성 데이터 집합에 대한 실험을 수행하였다. 제안한 방법은 지속적으로 순차적 스캔 방법보다 우수한 성능을 보였고, 성능이득은 대용량 비디오 데이타베이스에서도 유지되었다.

Keywords

References

  1. E. Ardizzone, M. La Cascia, A. Avanzato, and A. Bruna, 'Video Indexing Using MPEG Motion Compensation Vectors,' Proc. IEEE Int. Conf. on multimedia Computing System, 1999 https://doi.org/10.1109/MMCS.1999.778574
  2. S. Dagtas and A. Ghafoor, 'Indexing and Retrieval of Video based on Spatial Relation Sequences,' Proc. of ACM multimedia 1999, 1999 https://doi.org/10.1145/319878.319910
  3. A. Hampapur and R. bolle, 'Feature Based Indexing for Media Tracking,' Proc. Int'l Conf. on Multimedia and Expo, pp.67-70, Aug. 2000
  4. V. Kobla, D.S. Doermann, K-I. Kin, and C. Flaoutsos, 'Compressed domain video indexing techniques using DCT and motion vector information in MPEG video,' Proc. SPIE Conf. on Storage and Retrieval for Image and Video Databases, 1997 https://doi.org/10.1117/12.263408
  5. J. Meng and S. -F Chang, 'CVEPS - A Compressed Video Editing and Parsing System,' Proc. of ACM Multimedia 1996, 1996 https://doi.org/10.1145/244130.244145
  6. Emile Sahouria, and Avideh Zakhor, 'Motion Indexing of Video,' Proc. International Conference on Image Processing, 1997 https://doi.org/10.1109/ICIP.1997.638824
  7. Roland Tusch, Harald, Kosch, and Laszlo Boszormeny, 'VIDEX: An Integrated Generic Video Indexing Approach,' Proc. of ACM Multimedia 2000, 2000 https://doi.org/10.1145/354384.378988
  8. J. Wei, Z.N. Li, and I. Gertner, 'A novel motion-based active video indexing method,' Proc. IEEE int. Conf. on Multimedia Computing System, 1999 https://doi.org/10.1109/MMCS.1999.778502
  9. R. Mohan, 'Video Sequence Matching,' Proc. Int'l Conf. on Audio, Speech and signal processing, 1998
  10. A. Vailaya, W. Xiong, and A. K. Jain, 'Querty by Video Clip,' Proc. Int'l Conf. on Pattern Recognition, pp.909-911, 1998 https://doi.org/10.1109/ICPR.1998.711299
  11. R. Lienhart, C. Kuhmunch, and W. Effelsberg, 'On the Detection and REcognition of Television Comercials,' Proc. IEEE Conf. on Multimedia Computing and Systems, 1997 https://doi.org/10.1109/MMCS.1997.609763
  12. J. M. Sanchez, X. Binefa, J. Vitria, and P. Radeva, 'Local Color Analysis for Scene Break Detection Applied to TV commercials Recognition,' Proc. Visual 99, pp.237-244, June 1999
  13. D. A. Adjeroh, M. C. Lee, and I. King, 'A Distance Measure for Video Sequence Similarity Matching,' Proc. Int'l Workshop on Multi-Media Database Management Systems, 1998 https://doi.org/10.1109/MMDBMS.1998.709503
  14. D. M. Squire, H. Muller, and W. muller, 'Improving Response Time by Search Pruning in Content Based Image Retrieval System, Using Inverted File Techniques,' Proc. IEEE Workshop on Content Based Image and Video Libraries, pp.45-50, 1990 https://doi.org/10.1109/IVL.1999.781122
  15. G. A. Stephen, String Searching Algorithms, world Scientific Publishing, 1994
  16. A. Hampapur, K. Hyun and R. Bolle, 'Comparison of Sequence Matching Techinques for Video Copy Detection,' Proc. SPIE Conf. on storage and Retrieval for Media Databases, 2002.(accepted to appear) https://doi.org/10.1117/12.451091
  17. M Ellis, Method and systems for recognition of broadcast segments. In US5504518, United State Patent Office, 1996
  18. M. Ioka and M. Kurokawa, 'A Method for Retrieving Sequences of Images on the basis of Motion Analysis,' Proc. SPIE Conf. on Image Storage and Retrieval Systems, 1994 https://doi.org/10.1117/12.58490
  19. D. N. Bhat and S. K. Nayar, 'Ordinal Measures for Image Correspondence,' IEEE Trans. on PAMI, 20(4), pp. 415-423, 1998 https://doi.org/10.1109/34.677275
  20. http://www.dlib.org/dlib/february97/columbia/02chang.html#art