On a Hierarchical Indexing Fuzzy Content-Based Image Retrieval Approach.
Krisztián Veréb · Very Large Data Bases · 2003
In this paper a two-layer content-based image retrieval model will be described. The model has a prefilter indexing function, based on a possible class hierarchy of the images to be stored. Based on this hierarchical indexing some typified queries can be introduced. The δ distance function, used in the queries can be defined by a fuzzy algorithm based on the cut operation and the fuzzy logical connectives. This aprroach is called cut-and-or-not.