Algorithm to discriminate between new and used bills using a neural network

In this paper, we have proposed a method of classification for new and used currency bills from observed data using a neural network and a 2-stage adaptive digital filter (ADF). The 2-stage ADF is used in order to detect the desired signal from noisy input data. Whereas the neural network (NN) is used in order to discriminate new and used currency bill. The technique used here is based on an extension of the usual ADF designed by B. Widrow (1985) and the back-error-propagation (BP) method.