A SURVEY ON EVALUATING NEURAL NETWORK AND HIDDEN MARKOV MODEL CLASSIFIERS FOR HANDWRITING WORD RECOGNITION

Handwritten recognition is of immense importance for processing of bank checks, postal address, forms, mail or technical document. The recognition by the machine is difficult due to high variability and uncertainty of human writing. Handwritten words are fairly complex in pattern, great variability in handwriting style and in the character shapes by individuals. Neural Network and Hidden Markov Model is base for many different types of applications in various fields, many of which we use in our daily lives. The widely used methods are Neural Networks (NN) and Hidden Markov Models (HMM). NN classifier is used to generate a score for each segmented character. HMM classifier is used to identify character in sequence of word with assigning a probability to each of them. The main objective of this paper is to study various methods of Neural Networks (NN) and Hidden Markov Models (HMM) classifier applied to the handwritten word recognition problem. Exploring the result obtained from individual classifier, merits, demerits and the tradeoff of both classifiers in improving the throughput of handwritten word recognition system.