HOG-Based Feature Extraction for Accurate Modi Script Character Recognition

Modi is an old script used for documentation in Maharashtra that dates to the 17th century. With time the writing style of the script changes, and a lot of variation occurred till date. In the 20th century, due to a lack of awareness and understanding about the Modi-script, it disappeared and was replaced by the Devanagari. Manuscripts with Modi-script have not yet been explored, the information present in it remains unrevealed. To unfold the knowledge and develop a HOCR method for recognizing Modi characters, first need to understand the structural features of Modi characters. The main aim of this paper is to explore the structural features of the Modi characters that help in extracting the features.