Sub-structure Learning Based Handwritten Chinese Text Recognition

Yuanping Zhu,Jun Sun,Satoshi Naoi
DOI: https://doi.org/10.1109/ICDAR.2013.66
2013-01-01
Abstract:This paper proposed a sub-structure learning based method for handwritten Chinese text recognition. In conventional methods, a standard character recognizer is trained on character classes only. Unreliable recognition results on character segments will decrease final recognition precision. By discovering stable sub-structure patterns from real character segment samples automatically, both character and sub-structure patterns are trained in character recognizer. The judgment reliability of segments being characters is significantly improved. Furthermore, to deal with millions of training segment samples, a two-stage clustering method is proposed for sub-structure learning. Experiment results on HIT-MW database show that the sub-structure learning based method improves performance significantly. The F1-measure evaluation of handwritten Chinese text recognition is improved by 8.84%.
What problem does this paper attempt to address?