An wavelet-fractal neural network used in cutting tools wear monitoring

A cutting tool shape-recognizing algorithm based on the combination of wavelet-fractal feature extraction and neural network classification are proposed here. Image information of knife-point is first sent into edge detection model realized by wavelet multi-scale analyzing and transformed to edge configuration information. Then the fractal dimensions of the edge image in different resolution are computed to describe the local characteristic of the image outline. At last, all the feature parameters are sent into neural network classifier to get the wear state of cutting tools. The sub-systems are integrated together to realize the classification automatically and adaptively.