Study of off-line handwritten Chinese character recognition based on dynamic pruned FSVMs

According to the off-line handwritten Chinese characters, a classification and recognition method which is combined by pruning FSVM coarse classification and SVM fine classification is proposed in this text. First cut no value minor to reduce the number of support vector machines, and then determine the coarse classification through fuzzy membership when the coarse classification is done. In fine classification, OAA SVM algorithm is used to achieve the same Chinese characters recognition. The simulation result shows that this method can improve the recognition rate and speed of off-line handwritten Chinese.