GENERAL IMAGE RETRIEVAL MODEL
Peter L. Stanchev · 1998
In this paper we analyse the existing approaches to image retrieval and we propose a General Image Retrieval (GIR) model. The model establishes a taxonomy based on the systematisation on the existing approaches. In it the queries for image retrieval can be specified visually and interactively. The query mechanism is based on the calculation of similarities between the image query representation and the images in the image database. The GIR model provides numerous retrieval mechanisms such as: retrieval by attribute values, shape, colour, texture, visual examples, spatial similarity, and browsing and is general enough to serve as an engine to various image applications. It is designed to be used as a part of an Image Database System. An example for applying the GIR model to a plant image database is given.