Using a Haar wavelet transform, principal component analysis and neural networks for OCR in the presence of impulse noise

Abstract

In this paper we propose a novel algorithm for optical character recognition in the presence of impulse noise by applying a wavelet transform, principal component analysis, and neural networks. In the proposed algorithm, the Haar wavelet transform is used for low frequency components allocation, noise elimination and feature extraction. The principal component analysis is used to reduce the dimension of the extracted features. We use a set of different multi-layer neural networks as classifiers for each character; the inputs are represented by a reduced set of features. One of the key features of the proposed approach is creating a separate neural network for each type of character. The experimental results show that the proposed algorithm can effectively recognize the characters in images in the presence of impulse noise; the results are comparable with ABBYY FineReader and Tesseract OCR.

Minimisation of robust estimates of the sums of parametrised functions

“…For example, these approaches made possible application of weighted variants of back propagation algorithms for NN robust learning. Construction of robust learning algorithms of NN are important in a sense of many applications [12][13][14][15]. In particular an iteratively reweighted procedures are proposed.…”

A numerical approach to solve the model of an electromechanical system

“…For more information about Meyer wavelet, see other studies(); Daubechies wavelet, see other studies(); Fast wavelet see other studies(); and Haar wavelet, see other studies. ()…”

“…For more information about Meyer wavelet, see other studies 1-6 ; Daubechies wavelet, see other studies 7-10 ; Fast wavelet see other studies 11,12 ; and Haar wavelet, see other studies. [13][14][15] Haar wavelet family for t ∈ [0, 1) is defined by…”

Combined use of correlation measures for selecting semantically close concepts of the ontology

“…Traditionally, dimensional reduction techniques [11] were developed for the analysis of either quantitative (principal component analysis) [12] or categorical data (correspondence analysis). Lately a lot of attention has been paid to approaches to the analysis of discrete data.…”