Automatic Image Annotation (AIA)

Abstract

Image annotation an important research topic in the image retrieval field, where it is used to assign a specific words to image that describing their content. However, manual image annotation is no longer effective because it is tired and costly, so automatic image annotation becomes an alternative. In this thesis, we propose an approach for automatic image annotation based on probabilities. For a given concept, we associate the concept with a set of similar images. Then, we extracted a set of features from training images, thereafter we applied the Gaussian Mixture Model (GMM) to model the concept. We assign each image with 5 labels. The experimental results showed the effectiveness of the proposed approach as well as its superiority over other methods