A Spatial Similarity Ranking Framework for Symbolic Pictures Retrieval

A general spatial similarity ranking framework between two symbolic images with multiple pairwise spatial relations is proposed. The degree of similarity between two spatial relations is mapped to the distance between the associated nodes in an Interval Neighbor Group. The shorter the distance, the higher degree of similarity, while a longer one, a lower degree of similarity. Once all the pairwise similarity values are derived, an ensemble similarity will integrate these pairwise similarity assessments and give an overall similarity value between two images. Therefore, images in a database can be quantitatively ranked according to the degree of ensemble similarity with the query image. Two search strategies are discussed, namely, global and subimage similarity retrieval. A global strategy evaluates the ensemble similarity based on all pairwise spatial relations present in both images. On the other hand, only the spatial relations present in the maximum common subimage between the query and a database image are considered in the subimage strategy.