Artificial Neural Network based Classification of IC through Extracting the Feature Set of IC Images using 2-Dimensional Discrete Wavelet Transform

摘要

Machine Vision has rapidly increased its capabilities in the last few decades and Artificial Neural Networks (ANN) have played an important part in doing so. ANN's have been shown to be very effective tools for classification and sorting of data. In this paper, we use artificial neural networks and discrete wavelet (DWT) techniques for IC pin analysis and classification. Features are extracted from the wavelet transformed IC pin images and then fed into the ANN. The paper is limited to using two dimensional discrete wavelet transform and perceptron/ back-propagation algorithms of ANN.