A semantic approach for automatic image annotation

Abstract

Many features extraction method and classifiers are used singly, with modest results, for automatic image annotation. In order to improve the semantic image annotation accuracy, this document provides an automatic system to annotate image content by using a fusion of 3 classifier and a combination of some features extraction methods; multiclass support vector machine, multilayer neural network and nearest neighbour classifiers are combined together in order to classify and to find the appropriate keywords for this content. The color histograms and moments are used in this paper as features to represent image content. We support our case by experimental results obtained from two image databases (ETH-80 and coil-100 databases ).