As satellite images are widely used in a large number of applications in recent years, content-based image retrieval\r\ntechnique has become important tools for image exploration and information mining; however, their performances\r\nare limited by the semantic gap between low-level features and high-level concepts. To narrow this semantic gap,\r\na region-level semantic mining approach is proposed in this article. Because it is easier for users to understand\r\nimage content by region, images are segmented into several parts using an improved segmentation algorithm,\r\neach with homogeneous spectral and textural characteristics, and then a uniform region-based representation for\r\neach image is built. Once the probabilistic relationship among image, region, and hidden semantic is constructed,\r\nthe Expectation Maximization method can be applied to mine the hidden semantic. We implement this approach\r\non a dataset consisting of thousands of satellite images and obtain a high retrieval precision, as demonstrated\r\nthrough experiments.
Loading....