Since handwritten Chinese characters' style is uncertain and differs as people differ, this article brings forward a new Average Fuzzy Direction Code based on weight, so as to conquer traditional fuzzy arithmetic's shortcoming that is without enough generalization capability; At the same time it improves association rules in data mining and applies it to the process of handwritten Chinese characters' generalization and picking up of their abstract attributes. Thus exact denotation of handwritten Chinese characters is resolved, and simultaneously it picks up the handwritten Chinese characters' characteristic and mines improved association rules, and further achieves the purpose of handwritten Chinese characters' quick recognition. Therefore it resolves traditional pattern identification's problem of poor adaptability.
Citation:
Weiping Zhu, Wei Liu, Zhuqing Huang, "Average Fuzzy Direction Based Handwritten Chinese Characters Recognition Approach," wkdd, pp.42-47, First International Workshop on Knowledge Discovery and Data Mining (WKDD 2008), 2008