On-line recognition of cursive Korean characters using graph representation

The automatic recognition of cursive Korean characters is a difficult probl em, not only due to the multiple possible variations involved in the shapes of characters. but also because of the interconnections of neighboring gra phemes within an individual character. This paper proposes a recognition me thod for Korean characters using graph representation. This method uses a t ime-delay neural network (TDNN) and graph-algorithmic post-processor for gr apheme recognition and character composition, respectively. The proposed me thod was evaluated using multi-writer cursive characters in a boxed input m ode. For a test data set containing 26,500 hand-written cursive characters, a 92.3% recognition rate was obtained. (C) 2000 Pattern Recognition Societ y. Published by Elsevier Science Ltd. All rights reserved.