Manuscript Title:

TEXT LINE SEGMENTATION IN TAMIL LANGUAGE PALM LEAF MANUSCRIPTS – A NOVEL APPROACH

Author:

Dr. M. Mohamed Sathik, R. Spurgen Ratheash

DOI Number:

DOI:10.17605/OSF.IO/8DWSQ

Published : 2021-04-10

About the author(s)

1. Dr. M. Mohamed Sathik - Sadakathullah Appa College, Tirunelveli, Tamil Nadu, India, Manonmaniam Sundaranar University, Tamil Nadu, India.
2. R. Spurgen Ratheash - Sadakathullah Appa College, Tirunelveli, Tamil Nadu, India, Manonmaniam Sundaranar University, Tamil Nadu, India.

Full Text : PDF

Abstract

Segmentation of text lines from palm leaf manuscripts is an essential prior activity for character recognition. The scribers writing style creates intricacy in text line segmentation by low space between text lines and elongated characters placed in the text lines. Inefficient text line segmentation makes unproductive when promoting to character segmentation and character recognition process. The researchers have proposed a new way of text line segmentation algorithm named as Text Line Slicing algorithm for Tamil palm leaf manuscripts. This article explores text line segmentation from the scratch of preprocessing. The identification, segmentation of touching and overlapping text lines by an elongation of the character proves uniqueness of an algorithm. Text Line Slicing provides successful result in Tamil text line segmentation amidst several challenges. This outcome is an evidence of novelty among aplenty of text line segmentation methods in Tamil and other language palm leaf manuscripts.


Keywords

Binarization, Line segmentation, Obstacle, Palm leaf, Preprocessing, Tamil manuscripts, Text line slicing, Touching line, Overlapping lines