Online Handwriting Recognition for Tamil Script in Recent Past | Original Article

ABSTRACT:

The recent past has observed a rigorous research on optical character recognition and handwriting recognition (HWR) for Chinese-Japanese- Korean, Latin, Arabic and Indic scripts. Handwriting recognition can understand a handwritten text whether it is written on the paper, palm leaves, stone or digital surfaces. In case of offline handwriting recognition, the text is first written on the paper or any other hard surface, then it is scanned to convert it to digital form of text and finally the computer understands the handwritten text. In case of online handwriting recognition mode of HWR, the handwritten characterswordssentences are understood by the computer while writing on the digital surface of pen tablets, mobiles, computers or handwriting boards. An online handwritten word recognition (OHWR) model has various steps as data collection, preprocessing, segmentation of connected strokes, feature extraction, classification and post processing etc. The present work has been carried out for the Indic script Tamil, which is an abugida script and used Tamils and Tamil speakers in India. The present work presents the recent work done for Tamil handwriting recognition in online handwriting mode. This work has been done for handwriting recognition of Tamil characters and words. The major results represented in this paper have been taken from reputed journals and conferences of pattern recognition, artificial intelligence and handwriting recognition as Pattern Recognition, Pattern Analysis and Machine Intelligence, Pattern Recognition Letters, International Workshop on Frontiers in Handwriting Recognition, International Journal on Document Analysis and Recognition , International Conference on Document Analysis and Recognition and International Conference on Frontiers in Handwriting Recognition etc.