Please use this identifier to cite or link to this item:
Title: Context-aware features and robust image representations
Authors: Martins, P. 
Carvalho, P. 
Gatta, C. 
Keywords: Local features; Keypoint extraction; Image content descriptors; Image representation; Visual saliency; Information theory; Kernel estimators; Complementarity
Issue Date: Feb-2014
Publisher: Elsevier
Citation: MARTINS, P.; CARVALHO, P.; GATTA, C. - Context-aware features and robust image representations. "Journal of Visual Communication and Image Representation". ISSN 1047-3203. Vol. 25 Nº. 2 (2014) p. 339-348
Serial title, monograph or event: Journal of Visual Communication and Image Representation
Volume: 25
Issue: 2
Abstract: Local image features are often used to efficiently represent image content. The limited number of types of features that a local feature extractor responds to might be insufficient to provide a robust image representation. To overcome this limitation, we propose a context-aware feature extraction formulated under an information theoretic framework. The algorithm does not respond to a specific type of features; the idea is to retrieve complementary features which are relevant within the image context. We empirically validate the method by investigating the repeatability, the completeness, and the complementarity of context-aware features on standard benchmarks. In a comparison with strictly local features, we show that our context-aware features produce more robust image representations. Furthermore, we study the complementarity between strictly local features and context-aware ones to produce an even more robust representation.
ISSN: 1047-3203
DOI: 10.1016/j.jvcir.2013.10.006
Rights: openAccess
Appears in Collections:I&D CISUC - Artigos em Revistas Internacionais

Files in This Item:
Show full item record


checked on Nov 11, 2022

Citations 10

checked on Apr 2, 2022

Page view(s) 5

checked on Nov 28, 2022


checked on Nov 28, 2022

Google ScholarTM




Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.