A HANDWRITTEN CHINESE CHARACTER-RECOGNITION SYSTEM USING HIERARCHICALDISPLACEMENT EXTRACTION BASED ON DIRECTIONAL FEATURES

A recognition system using displacement extraction based on directiona l features is proposed for handwritten Chinese characters. In the syst em, after extracting the features from an input image, the displacemen t is extracted by the minimization of an energy functional, which cons ists of the Euclidean distance and the smoothness of the extracted dis placement. The coarse-to-fine strategy is adopted to escape local mini ma and reduce computational costs. The statistical classification is p erformed based on the estimated variance. In addition, the smoothness of the extracted displacement is utilized. An improvement in recogniti on performance is achieved as compared with the method without displac ement extraction. (C) 1998 Elsevier Science B.V. All rights reserved.