SEGMENTATION OF HANDWRITING NUMERICAL STRINGS IN FORM
Emil L. Co · 2003
Handwriting numeral segment is a hard but important task in an OCR system. In many applications, numerals are filled in preprinted form frames, which make the segmentation problem easier. General segment methods are developed for unconstrained numeral strings, but they are complex and slow. In this paper, a certain segment method for handwriting numeral strings in form frames is proposed. Usually, there are two connected types for numerals in form frames: transition-connected type (connect by a long horizon stroke) and share-connected type ( connect by sharing one period stroke). For the former type, we firstly detect the candidate segment positions based on local contour features, and then select a better position according to a 2-categories classifier. For the share-connected type, we segment by analysis the contour features. Experiment results on 574 real-life bank check images demonstrate the efficient of our method. The recognition success rate is 82. 4%, and the segment time is 13. 8ms per check. This method has been used in Automatic Chinese Bankcheck Processing demo system.