1 October 1998 Image retrieval based on wavelet vector quantization
Author Affiliations +
Abstract
In this paper, we describe a new image indexing and retrieval algorithm for large image databases based on the wavelets decomposition and vector quantization (VQ). The algorithm characterizes the color variations over the spatial extend of the image in a manner that provides semantically-meaningful image comparisons. To speed up retrieval, a two-step procedure is used that first makes a rough comparison based on the coarse features, and then refine the search by performing a fine feature vectors match between the selected images and the query. By adopting these wavelet VQ coding features, images can be compressed and indexed simultaneously, thus decreasing the complexity of database management. For the feasibility and practicality of the approach, a prototype system has been developed and tested with some experiments. Promising results have been obtained in experiments using a database of 15,000 general purposed images.
© (1998) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Tao Xia, Tao Xia, Jingli Zhou, Jingli Zhou, Shengsheng Yu, Shengsheng Yu, Rongfeng Yu, Rongfeng Yu, } "Image retrieval based on wavelet vector quantization", Proc. SPIE 3460, Applications of Digital Image Processing XXI, (1 October 1998); doi: 10.1117/12.323235; https://doi.org/10.1117/12.323235
PROCEEDINGS
11 PAGES


SHARE
Back to Top