A Novel Image Retrieval Approach Combining Multiple Features of Color-Connected Regions
This paper proposes a novel image retrieval method MCM (Multi-component Co-occurrence Matrices), which combines the color-connected regions in an image with their corresponding visual features in a region-growing like manner. Experimental results have shown that the MCM method has good retrieval performance and efficiency, due to the capability of integrating color composition, color spatial layout and texture characteristics into its coarse-granule region representation.
KeywordsImage Retrieval Query Image Dominant Color CBIR System Mosaic Block
Unable to display preview. Download preview PDF.
- 1.Malki, J., Boujemaa, N., Nastar, C., Winter, A.: Region queries without segmentation for image retrieval by content. In: Proceedings of Visual Information and Information Systems, Amsterdam, The Netherlands, June 2-4, pp. 115–122 (1999)Google Scholar
- 5.Vassili, K., Stephan, V.: Color co-occurrence descriptors for querying-by-example. In: Proceedings of International Conference on Multimedia Modeling, Lausanne, Switzerland, October 12-15, pp. 32–37 (1998)Google Scholar
- 6.Takahashi, N., Iwasaki, M., Kunieda, T., et al.: Image retrieval using spatial intensity features. Signal Processing: Image Communication 16, 45–57 (2000)Google Scholar
- 7.Yang, Y.B.: Research and applications on the key techniques of content-based image retrieval. Ph.D. Thesis, Nanjing University, Nanjing, P.R. China (in Chinese) (2003)Google Scholar