A fuzzy approach to content-based image retrieval

In this paper we present a fuzzy approach to content-based retrieval of segmented images. Our proposed methodology uses fuzzy attributed relational graphs (FARGs) and an efficient fuzzy graph matching algorithm (FGM). The graph matching algorithm is of low computational complexity and hence is suitable for content-based retrieval. Experimental results on random (sub)graph matching and content-based retrieval applications are also provided. The results indicate that the FARG-FGM framework of representing and retrieving images is efficient and accurate