Handwritten Character Recognition for Indian Languages

Abstract

Optical Character Recognition (OCR) is a crucial part of electronic processing and newlineanalysis of the documents called Electronic Document Analysis Systems. This area lies in the newlineintersection of the fields of Image Processing, Pattern Recognition, and Natural Language newlineProcessing. Although there have been tremendous research efforts, the state of the art in the newlineOCR has only reached the point of partial use in recent years. Nowadays, cleanly printed text newlinein documents with simple layouts can be recognized reliably by off the-shelf OCR software. newlineThere is only limited success in handwriting character recognition (HCR), and that too for newlineisolated and neatly hand-printed characters and words for limited vocabulary. However, in newlinespite of the intensive effort of more than thirty years, the recognition of free style and cursive newlinehandwriting continues to remain in the research arena. newlineIndia is a Multi-lingual country with many languages having their own distinctive newlinescripts and rich literary traditions. To help in bridging the digital divide and to increase the newlineease of communication with computers, it is particularly important to develop computer newlinesystems that allow users to interact with them in their respective languages. HCR can be used newlineas a solution to convert the hard copy information to soft form and also as an input media for newlinethe regional language scripts with large number of characters. The recognition of handwritten newlinecharacters for Indian languages is still an active research problem. newlineIn this thesis, we are investigating the challenges in building an offline HCR system newlinecapable of efficiently recognizing isolated handwritten characters including Kagunita for an newlineIndian script- Kannada. The aim is to investigate the HCR technologies that can be newlinegeneralized for various Indian language scripts. There is very little research base on HCR newlineavailable for Indian language scripts. We, therefore, surveyed national as well as international newlinelanguage OCRs including printed and handwritten scripts to find the suitable methodologies newlinefor each