Wavelet analysis for handwritten Chinese character recognition
Jun Yang, Songyu Yu, Zhao Rongchun · 2002
Wavelet transforms allow us to decompose a function into a multiresolution hierarchy of localized frequency bands, so we can simultaneously analyze characters' outline features at multiple levels of resolution. A new method for the recognition of handwritten Chinese characters based on wavelet transforms is introduced. We use the directional density vectors to preclassify the handwritten Chinese characters. The integration of the two methods provides a highly accurate and reliable recognition method.