Indic script family and its offline handwriting recognition for characters/digits and words: a comprehensive survey

Handwriting recognition has become an active area of research in pattern recognition and machine learning in recent years. Handwriting recognition systems have a variety of applications ranging from digital character conversion to signboard translation and scene image analysis. Extensive research work has been carried out for natural language processing of different scripts from all over the world, and the offline handwriting recognition of Indic scripts has also been drawing the major attention of pattern recognition researchers for the last two decades. The research for handwriting recognition has also gained further momentum because of planned funding by the Indian government towards technology development of Indian languages and scripts. These include work for Bangla, Devanagari, Gurmukhi, Tamil and Telugu, Gujarati, Kannada and other Indic scripts. In this study, we present an inclusive survey report of recent advances in Indic script offline handwriting recognition work, mainly focusing on the state-of-the-art work that appeared in the past 15 years, and it includes the following three related research areas in Indic scripts: (1) offline handwritten character recognition, (2) offline handwritten digit recognition and (3) offline handwritten word recognition. There is great work done for character and numeral hand-writing recognition in Indic scripts, but the research for word and sentence handwriting recognition in Indic scripts is still an emerging research area. Finally, the present study is an important guide for future researchers, and there exists a great opportunity for further research in Indic script offline handwriting recognition.