Character Segmentation in Machine Printed Malayalam Documents
Rajesh Gopakumar, U. Dinesh Acharya, N. V. Subba Reddy · 2005
Exclusive and intensive research has been done on optical character recognition and a large number of research papers have been published in the last few decades. Most of the efforts were made to develop OCR systems for foreign languages like English, Japanese, Roman and Arabic characters. Many commericial OCR systems for these foreign languages are avialable in the market. Even though in this direction efforts are made on some Indian Langauages, much owrk has not been done on any South indian languages, especially on Malayalm Script. Character Segmentation has been a critical area of OCR process. Segmentaton problem in Malayalam is a crucial problem as the characters are formed by combinaton of consonants, vowles and modifiers. In this paper we present a 2 stage character segmentation approach of which in the first coarse level stage, characters are segmentd with projection profile method and then in the second stage characters are segmented using zone level features and connected component analysis.