Best-match retrieval for structured images

This paper propose a new methodology for fast best-match retrieval of struc tured images. A triangle inequality property for the tree-distance introduc ed by Oflazer [17] is proven. This property is, in turn, applied to obtain a saturation algorithm of the trie used to store the database of the collec tion of pictures. The new approach can be considered as a substantial optim ization of Oflazer's technique and can be applied to the retrieval of homog eneous hierarchically structured objects of any kind. The new technique ins cribes itself in the number of distance-based search strategies and it is o f interest for the indexing and maintenance of large collections of histori cal and pictorial data. We demonstrate the proposed approach on an example and report data about the speed-up that it introduces in query processing. Direct comparison with MVP-trees algorithm is also presented.