A method for improving the machine recognition of confusing Chinese characters

This paper presents a novel low-cost method to improve the recognition of confusing Chinese characters which usually cause the misclassifications in character recognition systems. This proposed method can automatically analyze and select better discriminative features to differentiate confusing characters without any complex operations which are required in many other related approaches. According to our experimental results, this method improves the recognition rate of a recognition system by correcting about 30% of the misclassifications