A Color Image Retrieval Method Based on Local Histogram
In this paper, we shall propose a color image retrieval method. Among the methods of color image retrieval, the histogram technique only captures the global properties so that it cannot effectively characterize an image. To overcome this drawback, we propose a scheme to capture local properties so that it can do retrieval more accurately. In our method, we segment the original image into several subimage blocks, and we do color histogram with every subimage block. After combining color histogram vectors into a multi-dimensional vector, we use this multi-dimensional vector to search the database for similar images. The experimental results show that our method gives better performance than the original color histogram technique.
Unable to display preview. Download preview PDF.
- 2.Y. K. Chan, “Image Matching Using Run-Length Feature, ” Similar Image and Video Retrieval Systems Based on Spatial and Color Attributes, Ph. D. thesis, Department of Computer Science and Information Engineering, National Chung Cheng University, Republic of China, December 2000Google Scholar
- 5.D. S. Park, J. S. Park, T.Y. Kim and J. H. Han, “Image Indexing Using Weighted Color Histogram, ” Proceedings of International Conference on Image Analysis and Processing, 1999, pp. 909–914.Google Scholar
- 7.H. S. Sawhney and J. L. Hafner, “Efficient Color Histogram Indexing, ” Proceedings of IEEE International Conference on Image Processing, vol. 2, Austin, Texas, November 1994, pp. 66–70.Google Scholar
- 8.B. Shahrary, “Scene Change Detection and Content-Based Sampling of Video Sequence,” Proceedings IS&T/SPIE, Conference on Storage and Retrieval for Image and Video Databases, San Joe, CA, February 1995.Google Scholar