Embedding spatial information into image content description for scene retrieval

Rukoz, Marta; Hoang, Nguyen Vu; Gouet-Brunet, Valérie; Manouvrier, Maude

Rukoz, Marta; Hoang, Nguyen Vu; Gouet-Brunet, Valérie; Manouvrier, Maude (2010), Embedding spatial information into image content description for scene retrieval, Pattern Recognition, 43, 9, p. 3013-3024. http://dx.doi.org/10.1016/j.patcog.2010.03.024

Type

Article accepté pour publication ou publié

Date

2010

Journal name

Pattern Recognition

Volume

Number

Publisher

Elsevier

Pages

3013-3024

Abstract (EN)

This article presents Δ-TSR, an image content representation describing the spatial layout with triangular relationships of visual entities, which can be symbolic objects or low-level visual features. A semi-local implementation of Δ-TSR is also proposed, making the description robust to viewpoint changes. We evaluate Δ-TSR for image retrieval under the query-by-example paradigm, on contents represented with interest points in a bag-of-features model: it improves state-of-the-art techniques, in terms of retrieval quality as well as of execution time, and is scalable. Finally, its effectiveness is evaluated on a topical scenario dedicated to scene retrieval in datasets of city landmarks.

Subjects / Keywords

Spatial relationships; CBIR; Local image features; Scalability