Back to articles
Volume: 5 | Article ID: art00049
Combining color descriptors for improved codebook modelbased image retrieval
  DOI :  10.2352/CGIV.2010.5.1.art00049  Published OnlineJanuary 2010

In this paper we present the design of an image Content-Based Indexing and Retrieval (CBIR) system which, based upon existing implementations of a number of well-known color descriptors, makes use of the bag-of-words or codebook model in order to construct a robust approach to the retrieval of images from a database in a query-by-example context. A new object image database was constructed specifically for this task, in an attempt to challenge the invariance properties of the system under controlled conditions of illumination, point of view and scale. The system permits the combined use of up to two of the different color descriptors considered. The experiments run over a subset of the image database show an improvement of the obtained results under some of the tested combinations, as well as the effect of the variation of the employed codebook size.

Subject Areas :
Views 1
Downloads 0
 articleview.views 1
 articleview.downloads 0
  Cite this article 

Aitor Álvarez Gila, Guanqun Cao, Sheikh Faridul Hasan, Yu Hu, "Combining color descriptors for improved codebook modelbased image retrievalin Proc. IS&T CGIV 2010/MCS'10 5th European Conf. on Colour in Graphics, Imaging, and Vision 12th Int'l Symp. on Multispectral Colour Science,  2010,  pp 306 - 313,

 Copy citation
  Copyright statement 
Copyright © Society for Imaging Science and Technology 2010
Conference on Colour in Graphics, Imaging, and Vision
conf colour graph imag vis
Society of Imaging Science and Technology
7003 Kilworth Lane, Springfield, VA 22151, USA