Remote Sensing Image Retrieval Using a Context-Sensitive Bayesian Network with Relevance Feedback

In this paper, a novel relevance feedback is proposed to capture the user's query intention in large remote sensing databases. In order to increase the retrieval precision and reduce the retrieval time, the proposed relevance feedback approach firstly pre-selects a set of candidate images using semantic score functions as well as code co-occurrence matrices, which are de- rived from the context-sensitive Bayesian network. Since based on the user-supplied examples the semantic score function is able to measure the extent to which a given remote sensing image is related to a semantic concept, the pre-selected candidate images are highly related to the user's query intention and thus enable a better precision. Then, a computational more expensive region- based relevance feedback technique is carried out on the candi- date images. Since the number of candidate images is smaller by order of magnitude than the number of stored images, the re- trieval time is significantly reduced.