Information fusion techniques for automatic image annotation

Many recent techniques in Automatic Image Annotation use a description of image content based on visual symbolic elements associating textual labels through symbolic connection techniques. These symbolic visual elements, called visual terms, are obtained by a tokenization process starting from the values of features extracted from the training images data set. An interesting issue for this approach is to exploit, through information fusion, the representations with visual terms derived by different image features. We show techniques for the integration of visual information from different image features and compare the results achieved by them.