Empirical review on handwritten Devanagari script recognition techniques using AI approaches
Swati Jaiswal, Kartik Chaudhari, Suraj Tade, Shrushti Khirwadkar, Atharva Pande · International Journal of Creative Computing · 2024
The digitisation of handwritten documents plays a crucial role in facilitating access, editing, and long-term preservation of data. The Devanagari script, serving as a pillar for various Indian language scripts, presents a unique challenge due to the lack of official digitising tools. This study focuses on the comprehensive analysis of methods employed for recognising offline, handwritten Hindi characters, with the ultimate objective of developing a system that can accurately convert full phrases into digital form. While existing research predominantly concentrates on character-level recognition, this work covers all the key steps involved in word recognition, ranging from database processing to character identification and integration of recognised words into coherent sentences. By addressing each facet of Hindi character recognition, this study contributes to the advancement of digitisation techniques for the Devanagari script, enabling efficient access and manipulation of handwritten Hindi documents.