Distinction between Machine Printed Text and Handwritten Text in a Document

Dr. Rajesh Pathak, Ravi Kumar Tewari

Abstract: In many documents machine printed& handwritten texts are intermixed.Optical Character Recognition (OCR) techniques are different for machine printed and handwritten text, so it is necessary to separate these text before giving input to the OCR. In this paper we are proposing methodology for Hindi language. This methodology is based on structural features of text. Experimental results on a data set show that the proposed approach achieves an overall accuracy of 95.1%.

Keywords: OCR, Devanagari Script, Machine Printed Text, Hand written Text, Line Segmentation, Word Segmentation