Hand-written Numeral Recognition Based on Fuzzy C-means Algorithm

Fuzzy c-mean algorithm is sensitive to the initial value and its result is easy to fall into the partial minimum. Thus, two-stage fuzzy c-mean cluster algorithm is proposed. Firstly to estimate the classified number and the initial cluster center through the similar entropy(satisfies similarity and nearness), secondly carries on the cluster again through the fuzzy c-mean algorithm, finally use the two-stage Fuzzy C-Mean cluster recognition of the hand-written numerals based on the Zernike moments. The example given in the end of the paper testifies this method is effective and provides the theory for further establishment of the hand-written numeral recognition standard storehouse.