A STATISTICAL APPROACH TO TEXTURE IMAGE RETRIEVAL VIA ALPHA-STABLE MODELING OF WAVELET DECOMPOSITIONS
A STATISTICAL APPROACH TO TEXTURE IMAGE RETRIEVAL VIA ALPHA-STABLE MODELING OF WAVELET DECOMPOSITIONS
复制标题
DOI：
--
复制
发表时间：
1997
期刊：
影响因子：
--
通讯作者：
G. Tzagkarakis;P. Tsakalides
中科院分区：
文献类型：
--
作者：
G. Tzagkarakis;P. Tsakalides
This paper addresses issues that arise in content-based information retrieval systems, which employ statistical feature extraction and similarity measurements of texture images. First, we observe that statistical distributions with heavy algebraic tails, such as the alpha-stable family, are in many cases more accurate modeling tools for the wavelet coefficients of images than families with exponential tails, such as the generalized Gaussian. Motivated by our modeling results, we extend a new wavelet-based texture retrieval method introduced recently by Do and Vetterli by computing the Kullback-Leibler distance between alpha-stable distributions. We analyze the performance of the proposed retrieval method through experimental results on a database of texture images and we compare it to the performance of the traditional approaches.
