Automatic categorization of image regions using dominant color based vector quantization

Title: Automatic categorization of image regions using dominant color based vector quantization
Creator: Islam, Md; Zhang, Dengsheng; Lu, Guojun
Date: 2008
Type: Text; Conference paper
Identifier: http://researchonline.federation.edu.au/vital/access/HandleResolver/1959.17/70276
Identifier: vital:6577
Identifier: https://doi.org/10.1109/DICTA.2008.17
Identifier: ISBN:9780769534565
Abstract: This paper proposes a dominant color based vector quantization algorithm that automatically categorizes image regions. In contrast to the conventional vector quantization algorithm, the new algorithm effectively handles variable feature vectors like dominant color descriptors. Furthermore, the algorithm is guided by a novel splitting and stopping criterion which is specially designed for dominant color descriptors. This criterion helps the algorithm not only to learn the number of clusters, but also to avoid unnecessary over-fragmentations of region-clusters. Experimental result shows that the proposed approach categorizes image-regions with very high accuracy.
Publisher: Canberra, ACT IEEE Computer Society
Relation: Proceedings of the Digital Image Computing: Techniques and Applications p. 191-198
Rights: This metadata is freely available under a CCO license
Subject: 0801 Artificial Intelligence and Image Processing
Reviewed

		Thumbnail	File	Description	Size	Format