Generalized Hidden Markov Models To Handwritten Devanagari Word Recognition

Abstract - Hidden Markov Models (HMM) have long been a popular choice for Western cursive handwriting recognition following their success in speech recognition. Even for the recognition of Oriental scripts such as Chinese, Japanese and Korean, Hidden Markov Models are increasingly being used to model substrokes of characters. However, when it comes to Indic script recognition, the published work employing HMMs is limited, and generally focused on isolated character recognition. In this effort, a data-driven HMM-based handwritten word recognition system for Hindi, an Indic script, is proposed. Though Devanagari is the script for Hindi, which is the official language of India, its character and word recognition pose great challenges due to large variety of symbols and their proximity in appearance. The accuracies obtained ranged from 30% to 60% with lexicon. These initial results are promising and warrant further research in this direction. The results are also encouraging to explore possibilities for adopting the approach to other Indic scripts as well.