Xiubao Huang

Fabric defect detection based on multiple fractal features and support vector data description

Computer-vision-based automatic detection of fabric defects is one of the difficult one-class classification tasks in the real world. To overcome the incapacity of a single fractal feature in dealing with this task, multiple fractal features have been extracted in the light of the theory of and problems present in the box-counting method as well as the inherent characteristics of woven fabrics. Based on statistical learning theory, the up-to-date support vector data description (SVDD) is an excellent approach to the problem of one-class classification. A robust new scheme is presented in this paper for optimally selecting values of the parameters especially that of the scale parameter of the Gaussian kernel function involved in the training of the SVDD model. Satisfactory experimental results are finally achieved by jointly applying the extracted multiple fractal features and SVDD to the detection of defects from several datasets of fabric samples with different texture backgrounds.

Detection of Fabric Defects by Auto-Regressive Spectral Analysis and Support Vector Data Description

For the purpose of realizing fast and effective detection of defects in woven fabric, and in consideration of the inherent characteristics of fabric texture, i.e., periodicity and orientation, a new approach for fabric texture analysis, based on the modern spectral analysis of a time series rather than the classical spectral analysis of an image, is proposed in this paper. Traditionally, a power spectral estimated by a two-dimensional Fast Fourier transformation (FFT) is usually employed in the detection of fabric defects, which involves a large computational complexity and a relatively low accuracy of spectral estimation. To this effect, this paper makes a one-dimensional power spectral density (PSD) analysis of the fabric image via a Burg-algorithm-based Auto-Regressive (AR) spectral estimation model, and accordingly extracts features capable of effectively differentiating normal textures from defective ones. A support vector data description is adopted as a detector in order to deal with defect detection, a typical task of one-class classification. Experimental results for the detection of defects from several fabric collections with different texture backgrounds indicate that a low false alarm rate and a low missing rate can be simultaneously obtained with less computational complexity. Comparison of the detection results between the AR model and the FFT method confirms the superiority of the proposed method.

Modeling the air-jet flow field of a dual slot die in the melt blowing nonwoven process

An air-jet How field model of the dual slot die in the melt blowing nonwoven process is proposed and solved numerically with the finite difference method. The effects of dual slot die design parameters on velocity and temperature distributions in the melt blowing flow field are investigated. The computation results of the distributions of the x-compo nents of air velocity and air temperature coincide well with our experimental data. The results show that a smaller angle α, a larger slot width e, and a narrower head width f will all result in higher x-components of air velocity and higher air temperatures, which are beneficial to the air drawing of the polymer melt and thus to reduced fiber diameter. The results show the great potential of this research for computer-assisted design in the melt blowing nonwoven technology.

Effects of Processing Parameters on the Fiber Diameter of Melt Blown Nonwoven Fabrics

Our air drawing model of polymers in the melt blowing process is verified by the experimental results obtained with our university’s equipment. The predicted fiber diameters tally well with the experimental data. The effects of the processing parameters on the fiber diameters are further investigated in this paper. We find that a lower polymer flow rate, a higher initial polymer temperature, a higher initial air velocity, and a higher polymer melt flow index can all produce finer fibers, while the effect of the initial air temperature is insignificant. The results reveal the great potential of this research for computer assisted design of the melt blowing technology.

Predicting the fibre diameter of melt blown nonwovens: comparison of physical, statistical and artificial neural network models

Physical, statistical and artificial neural network (ANN) models are established for predicting the fibre diameter of melt blown nonwovens from the processing parameters. The results show that the ANN model yields a very accurate prediction (average error of 0.013%), and a reasonably good ANN model can be achieved with relatively few data points. Because the physical model is based on the inherent physical principles of the phenomena of interest, it can yield reasonably good prediction results when experimental data are not available and the entire physical procedure is of interest. This area of research has great potential in the field of computer assisted design in melt blowing technology.

A Practical and Robust Way to the Optimization of Parameters in RBF Kernel-Based One-Class Classification Support Vector Methods

Supported by one-sided samples information alone, one-class classification problems are more difficult to deal with than those of the traditional two-class or multi-class classification in the sense of parameters optimization. Support vector data description (SVDD) has become one of the most popular kernel learning methods for solving one-class classification problems, while RBF kernel is the most widely used kernel function. Though a good many researchers have jointly employed SVDD and RBF kernel, a rare of them discussed the parameters optimization in detail. Pointing out the deficiencies of the existing concerned approaches, this research proposed a new and practical way to the optimization of parameters in RBF kernel-based SVDD. Experimental results of textural defects detection validate the proposed method.