Efficient Segmentation Path Generation for Unconstrained Handwritten Hangul Character
This study suggests background thinning method for segmenting character unit of handwritten Hangul. Background thinning method conducts thinning processing using background information between characters and shows effective performance in segmenting for overlapped and touched characters. Character segmentation method using background thinning shows rapid segmentation performance with external segmentation which needs no judgment of recognition process. This method showed excellent performance in touched character segmentation as well as in segmentation of overlapped characters.
Unable to display preview. Download preview PDF.
- 1.Oliveira, L.S., Sabourin, R.: Automatic Recognition of Handwritten Numerical Strings: A Recognition and Verification Strategy. IEEE PAMI 24(11), 1438–1454 (2002)Google Scholar
- 2.Park, J.: An Adaptive Approach to Offline Handwritten Word Recognition. IEEE PAMI 24(7), 920–931 (2002)Google Scholar
- 4.Cho, S.-B., Kim, J.H.: A Hierarchical Organization of Neural Networks for Printed Hangul Character Recognition. Korea Information Science Journal 17(3), 306–316 (1990)Google Scholar
- 7.Kim, G., Yoon, J.J.: An Approach to Segmentation of Address Strings of Unconstrained Handwritten Hangul using Run-Length Code. Korea Information Science Journal 28(11), 813–821 (2001)Google Scholar
- 8.Choi, S.-M., Il-Seok Oh: Segmentation-free Recognition of Touching Numeral Pairs. Korea Information Science Journal 27(5), 563–574 (2000)Google Scholar
- 9.You, D., Kim, G.: An Efficient Slant Correction for Handwritten Hangul Strings using Structural Properties. Korea Information Science Journal 30(2), 93–102 (2003)Google Scholar
- 10.Gonzalez, R.C., Woods, R.E.: Digital Image Processing. Addison-Wesley Publishing Company, Reading (1992)Google Scholar
- 11.Park, J.-S., Hong, K.-C., Il-Seok Oh: Shape Decomposition of Handwritten hangul Characters. Korea Information Science Journal 28(7), 511–523 (2001)Google Scholar