A novel binarization technique based on Whale Optimization Algorithm for better restoration of palm leaf manuscript

  • PDF / 1,610,562 Bytes
  • 8 Pages / 595.276 x 790.866 pts Page_size
  • 75 Downloads / 209 Views

DOWNLOAD

REPORT


ORIGINAL RESEARCH

A novel binarization technique based on Whale Optimization Algorithm for better restoration of palm leaf manuscript T. Jerry Alexander1   · S. Suresh Kumar2 Received: 4 May 2020 / Accepted: 11 September 2020 © Springer-Verlag GmbH Germany, part of Springer Nature 2020

Abstract Palm leaf manuscript is a rich source of information pertaining to culture and heritage from the ancient period which is written on Palm leaves. They share surplus knowledge base in the field of art, medicine, culture, science, astronomy, astrology, crafts, literature, etc. Due to environmental effects and various other factors like handling, storage methods, etc. these manuscripts undergoes degradation. Hence for preservation of the information contained in the palm leaves, digitization of the manuscript is needed. The manuscript is either scanned or photographed for digitization and stored for preserving the content. For storage, enormous storage space is required and in addition, the degradation due to the noise should be removed for restoration of the text in the Palm leaf manuscript. Textual retrieval is done by using binarization method. Many binarization techniques such as Local threshold methods, Global thresholding, Otsu histogram thresholding, Adaptive thresholding techniques are available in the literature. Even then lots of challenges are present due to the degradation of the Palm-leaf manuscript. Swarm intelligence techniques are gaining importance in almost all domains for the optimization of parameters. In this present work, an attempt is made to use Whale Optimization Algorithm for optimization of Adaptive thresholding and the results are compared with the existing techniques. It proves the ascendance of the proposed technique over the prevailing techniques. Keywords  Binarization · Palm leaf manuscript · Niblack · Sauvola · Adaptive thresholding · Whale Optimization Algorithm · Text retrieval · Image restoration

1 Introduction Tamil is an ancient language used in South East Asia. For disseminating knowledge, information was presented in rocks, metals, leaves, cloth in old age even before the origin of the paper. In most parts of India, processed palm leaves were used. In Tamil Nadu, the palm manuscript is called OlaiChuvadi. Palm leaf manuscript, a popular way of writing is available since the fifth century B.C. It serves as a rich source for culture and heritage. Palm leaves: Palmyra and talipot are the commonly used source for the manuscript writing. Palm leaf manuscripts * T. Jerry Alexander [email protected] 1



Research Scholar, Faculty of Electronics Engineering, Sathyabama Institute of Science and Technology, Chennai, India



Principal, Swarnandhra College of Engineering and Technology, Narasapur, India

2

are normally in linear horizontal form due to leaf size with 15–60 cm length and 3–12 cm width. Young half-opened palm leaves are collected from the tree. Collected palm leaves are cut and boiled in water for a particular temperature to make them soft and later dried in shade followe