Hangul tree classifier for type clustering using horizontal and vertical strokes

Hangul is clustered into six different types in general. Type clustering has an effect of coarse classification in syllable matching and becomes the pre-processing stage for the segmentation in the grapheme matching process. We define a set of grapheme regions of preconsonants, vowel and post-consonant, that can absorb the change of the character's shape and the noise. A method for extracting the main stroke of the horizontal vowel and vertical vowel that appear in the region is proposed in this paper.