The huge amount of data as well as the complexity of individual objects are limiting factors for analysis and exploration of real data sets. These challenges are by developing new effective and efficient concepts for data analysis and data handling via data modelling. As a result, motivation for multi-resolution Image Similarity Modeling and a tendency towards multi-resolution image similarity modeling are at the core of the thesis. An exhaustive evaluation of the global, region-based and multi-resolution image similarity models is done on four different databases.