Neuro-classification of the new and used bills using time-series acoustic data

In this paper, we propose a neuro-classification method of the new and used bills using time-series acoustic data. The technique used here is based on an extension of an adaptive digital filter (ADF) by Widrow, individual adaptation (IA), and the error back-propagation (BP) algorithm. Two-stage ADF is used to detect the desired acoustic data(bill sound) of bill from noisy input(time-series acoustic) data. In the first stage, superfluous signals are eliminated from input signals and in the next stage, only the estimated(desired) acoustic data is detected from output signal of the two-stage ADF. The output signal of two-stage ADF is transformed into spectral data to generate an input pattern to a neural network (NN). The NN is used to discriminate the new and used bill. It is shown that in the experiment discrimination results using two-stage ADF are better than those obtained by using original observation data.