Journal of Computer Science and Technology

, Volume 17, Issue 6, pp 865–873 | Cite as

An image retrieval method using DCT features



In this paper a new image representation for compressed domain image retrieval and an image retrieval system are presented. To represent images compactly and hierarchically, multiple features such as color and texture features directly extracted from DCT coefficients are structurally organized using vector quantization. To train the codebook, a new Minimum Description Length vector quantization algorithm is used and it automatically decides the number of code words. To compare two images using the proposed representation, a new efficient similarity measure is designed. The new method is applied to an image database with 1,005 pictures. The results demonstrate that the method is better than two typical histogram methods and two DCT-based image retrieval methods.


image retrieval vector quantization color histogram DCT 


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. [1]
    Gudivada V N, Raghavan V W. Content based image retrieval systems.IEEE Computer, 1995, 28(9): 18–22.Google Scholar
  2. [2]
    Bach J R, Fuller C, Gupta Aet al. The virage image search engine: An open framework for image management. InProc. SPIE 2670: Storage and Retrival for Still Image and Video Databases IV, San Jose, CA, USA, Feb., 1996, pp.76–87.Google Scholar
  3. [3]
    Mandal M K, Idris F, Panchanathan S. A critical evaluation of image and video indexing techniques in the compressed domain.Image and Vision Computing, 1999, 17: 513–529.CrossRefGoogle Scholar
  4. [4]
    Smith J R, Chang S F. Transform features for texture classification and discrimination in large image databases. InProc. IEEE Int. Conf. Image Processing, Austin, Texas, 1994, (3): 407–411.Google Scholar
  5. [5]
    Reeves R, Kubik K, Osberger W. Texture characterization of compressed serial images using DCT coefficients. InProc. SPIE 3022: Storage and Retrieval for Image and Video Databases V,San Jose, California, 1997, pp.398–407.Google Scholar
  6. [6]
    Furht Borko, Saksobhavivat P. A fast content-based multimedia retrieval technique using compressed data. InSPIE 3527: Conference on Multimedia Storage and Archiving Systems III, Boston, 1998, pp.561–571.Google Scholar
  7. [7]
    Shneier M, Mottaleb M A. Exploiting the JPEG compression scheme for image retrieval.IEEE Transactions on Pattern Analysis and Machine Intelligence, 1996, 18(8): 849–853.CrossRefGoogle Scholar
  8. [8]
    Abdel-Malek A A, Hershey J E. Feature cueing in the discrete cosine domain.Journal of Electronic Imaging, 1994, (3): 71–80.CrossRefGoogle Scholar
  9. [9]
    Shen B, Sethi I K. Direct feature extraction from compressed images. InProc. SPIE, 2670, 1996, pp.404–414.Google Scholar
  10. [10]
    Chang R F, Kuo W J, Tsai H C. Image retrieval on uncompressed and compressed domains. InICIP 2000, Toronto, Canada, 2000.Google Scholar
  11. [11]
    Wallace G K. The JPEG still picture compression standard.Communication ACM, 1991, 34(4): 31–45.CrossRefGoogle Scholar
  12. [12]
    Richard E, Robert S Ledley. Texture discrimination using discrete cosine transformation shift-insensitive descriptorsPattern Recognition, 2000, 33: 1585–1598.CrossRefGoogle Scholar
  13. [13]
    Huang J, Kumar S R. Spatial color indexing and applications.International Journal of Computer Vision, 1999, 35(3): 245–268.CrossRefGoogle Scholar

Copyright information

© Science Press, Beijing China and Allerton Press Inc. 2002

Authors and Affiliations

  1. 1.ATR National LabNational University of Defense TechnologyChangshaP.R. China

Personalised recommendations