Relation Bag-of-Features for Symbol Retrieval - Inria - Institut national de recherche en sciences et technologies du numérique Access content directly
Conference Papers Year : 2013

Relation Bag-of-Features for Symbol Retrieval


In this paper, we address a new scheme for symbol retrieval based on relation bag-of-features (BOFs) which are computed between the extracted visual primitives. Our feature consists of pairwise spatial relations from all possible combina tions of individual visual primitives. The key characteristic of the overall process is to use topological information to guide directional relations. Consequently, directional relation matching takes place only with those candidates having similar topological configurations. A comprehensive study is made by using two different datasets. Experimental tests provide interesting results by establishing user-friendly symbol retrieval application.
Fichier principal
Vignette du fichier
kc_icdar2013_2.pdf (568.78 Ko) Télécharger le fichier
Origin : Files produced by the author(s)

Dates and versions

hal-00823960 , version 1 (20-05-2013)


  • HAL Id : hal-00823960 , version 1


Santosh K.C., Laurent Wendling, Bart Lamiroy. Relation Bag-of-Features for Symbol Retrieval. ICDAR - International Conference on Document Analysis and Recognition - 2013, Aug 2013, Washington DC, United States. ⟨hal-00823960⟩
141 View
156 Download


Gmail Facebook X LinkedIn More