A Graph Based Approach for Handwritten Devnagari Word Recognition

A graph based holistic approach for handwritten Devnagari OCR is presented as part of a knowledge-based word interpretation model. This new method is based on the recognition of sub graphs homeomorphism to previously defined prototypes of words. In the system, feature graph is extracted from input word. It is extended to net. Net is created by generating new nodes. Each sub graph that is recognized is introduced as a node in a directed net that compiles different alternatives of interpretation of features in the feature graph. A path in the net represents a consistent succession of characters in the word. A final search for optimal path under certain criteria gives the best interpretation of the word features. The word level accuracy is 84% on the test set. Further the accuracy can be increased with the use of lexical context to find a meaningful interpretation.