Fault diagnosis for industrial images using a min-max modular neural network

This paper presents a new fault diagnosis method for industrial images based on a Min-Max Modular (M-3) neural network and a Gaussian Zero-Crossing (GZC) function. The most important advantage of the proposed method over existing approaches such as radial-basis function network and support vector machines is that our classifier has locally tuned response characteristics and the misclassification rate of faulty product images can be controlled as small as needed by turning two parameters of the GZC function while the correct rate can be influenced to some extend. The experimental results on a real-world fault diagnosis problem of industrial images indicate that the effectiveness of the proposed method.