Retrieve of images by content
Chaabane Djéraba, Ian Savory, H. Briand · 2002
In this paper, we present an approach of image retrieval by content in an object oriented database. The content of image includes automatically extracted features such as colors, textures, and manually extracted features such as relevant regions, shapes, key words and texts. All these features and their behaviors are represented in an object-oriented database. Finally a KD-tree is used to model the inner of features, and to support efficiently the match engine.