Directional Discrete Cosine Transform for Handwritten Script Identification

Abstract : This paper presents directional discrete cosine transforms (D-DCT) based word level handwritten script identification. The conventional discrete cosine transform (DCT) emphasizes vertical and horizontal energies of an image and de-emphasizes directional edge information, which of course plays a significant role in shape analysis problem, in particular. Conventional DCT however, is not efficient in characterizing the images where directional edges are dominant. In this paper, we investigate two different methods to capture directional edge information, one by performing 1D-DCT along left and right diagonals of an image, and another by decomposing 2D-DCT coefficients in left and right diagonals. The mean and standard deviations of left and right diagonals of DCT coefficients are computed and are used for the classification of words using linear discriminant analysis (LDA) and K-nearest neighbour (K-NN). We validate the method over 9000 words belonging to six different scripts. The classification of words is performed at bi-scripts, tri-scripts and multi-scripts scenarios and accomplished the identification accuracies respectively as 96.95%, 96.42% and 85.77% in average.
Type de document :
Pré-publication, Document de travail
Authors' copy - ICDAR International Conference on Document Analysis and Recognition (2013), Washi.. 2013
Liste complète des métadonnées

Littérature citée [19 références]  Voir  Masquer  Télécharger

https://hal.inria.fr/hal-00826298
Contributeur : Santosh K.C. <>
Soumis le : lundi 27 mai 2013 - 19:09:05
Dernière modification le : mardi 24 avril 2018 - 13:29:58
Document(s) archivé(s) le : mardi 3 septembre 2013 - 09:45:37

Fichier

CRC_Final.pdf
Fichiers produits par l'(les) auteur(s)

Identifiants

  • HAL Id : hal-00826298, version 1

Collections

Citation

Mallikarjun Hangarge, Santosh K.C., Rajmohan Pardeshi. Directional Discrete Cosine Transform for Handwritten Script Identification. Authors' copy - ICDAR International Conference on Document Analysis and Recognition (2013), Washi.. 2013. 〈hal-00826298〉

Partager

Métriques

Consultations de la notice

301

Téléchargements de fichiers

635