Handwritten Chinese Similar Characters Recognition Based On AdaBoost

As an effective statistical learning tool, AdaBoost has been widely used in the field of pattern recognition. Considering the large categories of Chinese characters and the recognition rate is low for similar character set, we propose a two-stage booting based scheme for Chinese character recognition. The AdaBoost procedure is applied to the similar character sets' recognition after the traditional distance classifier. Experimental results show that AdaBoost is much better than conventional distance classifier.