Off-line Handwritten Word Recognition Using a Mixed HMM-MRF Approach
Résumé
In this paper we present a two-dimensional stochastic method for the recognition of unconstrained handwritten words in a small lexicon. The method is based on an efficient combination of hidden Markov models ({\sc hmm}s) and causal Markov random fields ({\sc mrf}s). It operates in a holistic manner, at the pixel level, on scaled binary word images which are assumed to be random field realizations. The state-related random fields act as smooth local estimators of specific writing strokes by merging conditional pixel probabilities along the columns of the image. The {\sc hmm} component of our model provides an optimal switching mechanism between sets of {\sc mrf} distributions in order to dynamically adapt to the features encountered during the left-to-right image scan. Experiments performed on a French omni-scriptor, omni-bank database of handwritten legal check amounts provided by the A2iA company are described in great extent.