Comparative Study of Feature Extraction Techniques for Handwritten Ancient Modi Script Recognition

Abstract

In Maharashtra, documentation has been carried out using Devanagari script since the 19th century. Before the adoption of Devanagari, the Modi script was predominantly used for documentation, but its cursive writing style, complexity, and limited awareness led to its decline. The Modi script has deep historical origins, and numerous manuscripts are stored in Museums, National Libraries, and cultural heritage sites, such as temples,forts. Unfortunately, much of the knowledge contained in these documents remains inaccessible due to the lack of proper resources. The Government is working on preserving these documents and revealing the information they contain. Utilization of modern technologies, such as HOCR (Handwritten Optical Character Recognition), is now underway to restore cultural heritage. This study specifically focuses on recognizing script data from image documents, particularly handwritten Modi script. Feature extraction techniques play a crucial role in identifying text from the ancient Modi script, as illustrated by the comparative analysis provided here.