Enhanced document image Classification technique using Extreme learning machine classifier With fuzzy unsharp masking scheme

Abstract

Document Image Analysis for information extraction has very newlineprominent in recent years Document image is composed of a variety of newlinephysical entities or regions such as handwritten text machine printed newlinetexts blocks lines words figures tables and background Document newlineanalysis and recognition aims to automatic extraction of information newlinepresented in scanned documents The most widely known applications newlineof document analysis and recognition are processing of office newlinedocuments business letters forms engineering drawings and maps text newlinebooks technical manual musical notations and symbolic data Major newlineissues in the document analysis are removal of noise and artifacts that newlineare introduced during scanning phase segmentation feature extraction newlineand classification newline newline