An Extensive Review on Recognition of Antique Tamil characters for Information Repossession from Epigraphic Inscriptions

Abstract

The ancient script recognition is one of the most demanding areas for the researches due to numerous variations in writing styles. All-embracing research work is reported on recognition of various ancient scripts such as Brahmi, Grantha. However the work accounted on Tamil scripts is still in its infancy so momentous research is required in this field. The Tamil scripts which contain loop, curves, circle, dots, etc are the complex issues for identifying the characters present in the image or the input data. The Tamil script has many characters with many features in which those features should be extracted in an efficient way to envisage the character with higher accuracy. There is no proper technology to convert the ancient inscriptions to human readable form in digitised format. Several scripts and inscriptions are examined, which indicates some future research prospects. Based on study demeanour it has been found that there is a need of an automated system for ancient character recognition with hybrid feature extraction and classification techniques for attained the accurate results. This paper aims to provide the insights of types of scripts, inscriptions, various methods, techniques of feature extraction, classification reported over the last few years in the field of epigraphy. So, a novel framework has been proposed to recognize the ancient Tamil character with higher classification accuracy with an aim to surpass the workflow of existing systems.