An efficient selected feature set for the middle age Persian character recognition

Author

Alirezaee, S. ; Aghaeinia, H. ; Ahmadi, M. ; Faez, K.

Author_Institution

Dept. of Electr. Eng., Amirkabir Univ. of Technol., Tehran, Iran

fYear

2004

fDate

13-15 Oct. 2004

Firstpage

246

Lastpage

250

Abstract

In this paper, a morphological based method for recognition of handwritten middle Persian characters is presented. After pre-processing and noise cancellation, morphological erosion operator with many structure elements is applied. The structure elements are with variable length lines at directions 0, 45, 90, 135 degrees. A five element feature set has been defined so: (1) relative energy of eroded version with respect to the original image energy (REL_ENG),(2) displacement of the center of mass (CM__DIS), (3) minimum eigenvalue (EIG_MIN), (4) maximum eigenvalue (EIG_MAX) and (5) its direction (EIG-DIR). These features are used to design a feedforward neural network with one hidden layer. The best classification error is about 2.39% (97.61% recognition rate), and is achieved with 150 neurons for the hidden layer.

Keywords

eigenvalues and eigenfunctions; feedforward neural nets; handwritten character recognition; image denoising; efficient selected feature set; feedforward neural network; handwritten middle age Persian character recognition; maximum eigenvalue; minimum eigenvalue; morphological erosion operator; noise cancellation; original image energy; relative energy; Character recognition; Data mining; Eigenvalues and eigenfunctions; Entropy; Feature extraction; Feedforward neural networks; Natural languages; Neural networks; Telecommunication computing; Text recognition;

fLanguage

English

Publisher

ieee

Conference_Titel

Information Theory, 2004. ISIT 2004. Proceedings. International Symposium on

ISSN

1550-5219

Print_ISBN

0-7695-2250-5

Type

conf

DOI

10.1109/AIPR.2004.12

Filename

1409706