A PCA based method for improving the reliability of bank note classifier machines

Author

Ahmadi, Ali ; Omatu, Sigeru ; Kosaka, Toshihisa

Author_Institution

Osaka Prefecture Univ., Sakai, Japan

Volume

1

fYear

2003

fDate

18-20 Sept. 2003

Firstpage

494

Abstract

This paper addresses the reliability of neuro-classifiers for bank note recognition. A local principal component analysis (PCA) method is applied to remove non-linear dependencies among variables and extract the main principal features of data. At first the data space is partitioned into regions by using a self-organizing map (SOM) clustering and then the PCA is performed in each region. A learning vector quantization (LVQ) network is employed as the main classifier of the system. By defining a new algorithm for rating the reliability and using a set of test data, we estimate the reliability of the system. The experimental results taken from 1,200 samples of US dollar bills show that the reliability is increased up to 100% when the number of regions as well as number of codebook vectors in the LVQ classifier is taken properly.

Keywords

bank data processing; document image processing; image classification; image coding; principal component analysis; reliability; self-organising feature maps; vector quantisation; PCA based method; bank note classifier machines; codebook vectors; learning vector quantization network; neuro-classifiers; principal component analysis method; self-organizing map clustering; Cepstrum; Clustering algorithms; Data mining; Feature extraction; Neural networks; Partitioning algorithms; Principal component analysis; Robustness; System testing; Vector quantization;

fLanguage

English

Publisher

ieee

Conference_Titel

Image and Signal Processing and Analysis, 2003. ISPA 2003. Proceedings of the 3rd International Symposium on

Print_ISBN

953-184-061-X

Type

conf

DOI

10.1109/ISPA.2003.1296947

Filename

1296947