Advanced annotation techniques of multimedia data significantly improve representing and retrieving multimedia-based contents. In this paper we present an intelligent framework for attaching semantic annotations to image contents based on the extraction of elementary low-level features, user's relevance feedback, and the usage of ontology knowledge. This approach facilitates image annotation by computing a Semantic Annotation Template (SAT) of most likely relevant content descriptors as a result of extracted low-level features and the prior annotation of similar images. In order to generate the annotation templates, a multi-level relevance feedback is used, which interactively refines the properties of user's high-level concepts on both visual and semantic level.