FReBIR : an image retrieval system based on fuzzy region matching

Sylvie Philipp-Foliguet (1) , Julien Gony (1) , Philippe-Henri Gosselin (2)

1 ETIS - UMR 8051 - Equipes Traitement de l'Information et Systèmes
2 MIDI - Multimedia Indexation and Data Integration

In this paper we present a method of image indexing and retrieval which takes into account the relative positions of the regions within the image. Indexing is based on a segmentation of the image into fuzzy regions ; we propose an algorithm which produces a fuzzy segmentation. The image retrieval is based on inexact graph matching, taking into account both the similarity between regions and the spatial relation between them. We propose, on one hand a solution to reduce the combinatorial complexity of the graph matching, and on the other hand, a measure of similarity between graphs allowing the result images ranking. A relevance feedback process based on region classifiers allows then a good generalization to a large variety of the regions. The method is adapted to partial queries, aiming for example at retrieving images containing a specific type of object. Applications may be of two types, firstly an on-line search from a partial query, with a relevance feedback aiming at interactively leading the search, and secondly an off-line learning of categories from a set of examples of the object. The name of the system is FReBIR for Fuzzy Region-Based Image Retrieval.

Mots clés: Image indexing Image retrieval Inexact graph matching Fuzzy segmentation Spatial relationship Kernel function

Domaines: Vision par ordinateur et reconnaissance de formes [cs.CV]

Sylvie Philipp-Foliguet, Julien Gony, Philippe-Henri Gosselin. FReBIR : an image retrieval system based on fuzzy region matching. Computer Vision and Image Understanding, 2009, 113, pp.693-707. ⟨hal-00519447⟩