SEMANTIC-BASED USER DEMAND MODELING FOR REMOTE SENSING IMAGES RETRIEVAL

This paper aims at providing a more convenient approach for remote sensing images retrieval based on sematic-based user demand modeling. The semantic user demand model is a two-layer model that bridges the gap between users' satellite image demand of natural language description and satellite images. Natural language process and semantic inference are involved to generate the semantic user demand model. A knowledge database consists of ontologies, rules and dictionaries is developed to support natural language process and semantic inference. Semantic similarity and confliction-resolution are also adopted in inference. Finally, the model is validated by a prototype system based on protege-owl and JESS. The results show that the model and the approach are available.