Principle component analysis (PCA) based coin-counting system

Abstract

In this paper, a neural network using a feature extraction scheme known as principle component analysis (PCA) is proposed to recognize two-dimensional objects in an image. This approach consists of two stages. First, the procedures of determining the coefficients of rapid descriptor (RD) of 2-D objects from their boundary are described. To speed up the learning process of the neural network, a PCA technique is used to extract the principal components of these RD coefficients. Then, these reduced components are utilized to train a feed-forward neural network for object recognition and classification. We compare recognition performance, network sizes, and training time for networks trained with both reduced and unreduced data. The experimental results show that a significant reduction in training time can be achieved without a sacrifice in classifier accuracy.