Generalizing Texture Images Using Wavelet-Based Image Retrieval Algorithms

Data generalization is a data preprocessing technique employed in data mining. It is a process that abstracts a large set of task-relevant data in a database from a relatively low conceptual level to higher conceptual levels with the aim of reducing data size. This paper discusses a novel method of generalizing texture images by clustering them based on color, texture and semantically classifying them. Wavelet-based image retrieval algorithms are used in clustering and classifying. Experimental results show a promising result: texture images could be generalized using these algorithms. However, more work is needed to find more efficient techniques and a system involving more users’ participation with the aim of generating more satisfying result.