Handwritten Character Recognition Using Wavelet Transform and Hopfield Network

An off-line handwritten alphabetical character recognition system using Wavelet Transform and Hop field Network is described in the paper. Hop field Network has been used in the system as they are known for their ability to retain patterns. The wavelet transforms are used for extracting features from the images. The results show that the network was able to recognize all the characters at a distortion level of 30%, at 40% it recognized only a few characters and at all the distortion levels above 40% it was not able to recognize any of the characters.