Améliorer la classification de documents par combinaison de descripteurs visuels et textuels

Abstract

The main contribution of this paper is a new method for classifying document images by combining textual and visual features repectively extracted with the Bag of Words (BoW) and the Bag of Visual Words (BoVW) techniques. While previous attempts have been showing disappointing results by combining visual and textual features with the Borda-count technique, we’re proposing here a combination through learning approach. The other contribution of this paper are the experiments conducted on a 1925 document image industrial database revealing that this fusion scheme significantly improves the classification performances. Our concluding contribution deals with the choosing and tuning BoW/BoVW techniques in an industrial context. MOTS-CLÉS : BoW, BoVW, combinaison texte image, classification, application industrielle

DOI: 10.24348/sdnri.2014.CIFED-9

Extracted Key Phrases

3 Figures and Tables

Cite this paper

@inproceedings{Augereau2014AmliorerLC, title={Am{\'e}liorer la classification de documents par combinaison de descripteurs visuels et textuels}, author={Olivier Augereau and Nicholas Journet and Jean-Philippe Domenger}, booktitle={CORIA-CIFED}, year={2014} }