Semantic Enrichment for Automatic Image Retrieval

This chapter explains the algorithms and mechanisms used for semantic image indexing and retrieval on the Internet. It aims to develop a technique which captures local color and texture descriptors in a coarse segmentation framework of grids, and has a shape descriptor in terms of invariant moments computed on the edge image. The number of web images is increasing at a rapid rate, and searching them semantically presents a significant challenge. The effectiveness of image retrieval depends on meaningful indexing; the key problem of image retrieval is to organize them based on semantics. Despite continuous research efforts in developing and exploring new models, the gap between the expressive power of image features and semantic concepts is still a fundamental barrier. The tagging process involves interpretation of the visual information given some context, either the context of the image or the context of the annotation or retrieval.