A Bayesian network for combining descriptors: application to symbol recognition - Inria - Institut national de recherche en sciences et technologies du numérique Access content directly
Journal Articles International Journal on Document Analysis and Recognition Year : 2009

A Bayesian network for combining descriptors: application to symbol recognition

Abstract

In this paper, we propose a descriptor combination method, which enables to improve significantly the recognition rate compared to the recognition rates obtained by each descriptor. This approach is based on a probabilistic graphical model. This model also enables to handle both discrete and continuous-valued variables. In fact, in order to improve the recognition rate, we have combined two kinds of features: discrete features (corresponding to shapes measures) and continuous features (corresponding to shape descriptors). In order to solve the dimensionality problem due to the large dimension of visual features, we have adapted a variable selection method. Experimental results, obtained in a supervised learning context, on noisy and occluded symbols, show the feasibility of the approach.
Fichier principal
Vignette du fichier
fulltext.pdf (391.37 Ko) Télécharger le fichier
Origin : Files produced by the author(s)
Loading...

Dates and versions

inria-00437492 , version 1 (30-11-2009)

Identifiers

Cite

Sabine Barrat, Salvatore Tabbone. A Bayesian network for combining descriptors: application to symbol recognition. International Journal on Document Analysis and Recognition, 2009, 13 (1), pp.65-75. ⟨10.1007/s10032-009-0103-y⟩. ⟨inria-00437492⟩
151 View
258 Download

Altmetric

Share

Gmail Facebook X LinkedIn More