Radial Basis Function for Visual Image Retrieval

Abstract

Visual image retrieval systems imply novel approaches to improve their performance. Search methods require mathematical formalism for feature extraction of multi medial data. Euclidean distance is not enough approach for extracting similarity between images. Radial basis function provides necessary optimization for similarity measures when query image is used as reference. This paper explains a novel method exploiting radial basis function applied for visual image retrieval area. The goal of the approach is the searching process optimization for image retrieval which is obtained results reveal interesting conclusions about advantages of the proposed method even conventional similarity measurement approaches used in content based retrieval area.