Query-by-shape in meteorological image archives using the point diffusion technique

In this paper, we work on meteorological satellite image archives and provi de a novel and useful query-by-shape tool. To this aim, we first present th e point diffusion technique (PDT), a fast and efficient method for shape si milarity evaluation. Thanks to its very structure, this approach is suitabl e to handle objects whose shape is not well defined and can be represented by a set of sparse points. PDT is thus suitable for application to similari ty-based retrieval from remotely sensed image archives, where shapes are ha rdly defined but are still among the major features of interest. Moreover, we prove here that PDT is almost as effective as more standard pr ocedures for shape-based database queries, although significantly faster. I n other words, it manages to combine retrieval speed and precision, the fea tures of greatest importance for a first remote sensing data prescreening i n many applications. Archives of meteorological satellite images are typica l examples of very large-sized, remote sensing-based databases with a speci al attention for shape features. Each meteorological satellite produces ter abytes of data every day, a large part of which is not immediately analyzed and ends being stored in archives. In this paper, the application of PDT t o such a database is presented and discussed, and a comparison with a stand ard method developed for meteorological shape analysis is provided.