International Journal of Electrical, Electronics and Computer Engineering (IJEECE)
In postal system most often the users find unreadable handwritten characters and numerals. In digital system also due to bad transmission some data becomes corrupted. So getting the original data from the corrupted data is an important issue. Hopfield network is one of the methods which choose the original data that best matches with the corrupted data from a pre-stored dataset. Here the author has proposed an advance dual neuron scheme for this purpose. This method employs simple hopfield network to store patterns and as a recognizer it choose two consecutive neurons at a time to steer the corrupted or partial data into the original data.