With the Kohonen network clustering in neural network employed, the degree of relationship of the universal joint axle of the rolling mill was input to Kohonen network as the training sample, studied and clustered by the network to generate different clustering centers according to the different depth and different degree of relationship among the cracks.

Simulation experiments were carried out on a test rig for single-span rotor system with crack fault and the collected fault signals were analyzed by way of a 3-D waterfall spectra in combination with reassigned wavelet scalogram.

The moire phenomena appeared when scanning a halftone image was theoretically explained.

In order to solve the moire caused by the superposition of several plates in the course of screen printing,the forming process of moiré was analyzed with mathematics method from the formation mechanism.

This paper describe the concept of moire occured in screen printing and the difference between the moire in screen printing and the moire in lithosraphy.

In real printing thereasily exists moiré because of regular gridding module.

Current resampling techniques computer new samples by simple interpolation methods like bilinear interpolation which can cause undesirable moiré artifacts when the original image contains high frequency components.