Nonwoven uniformity identification using wavelet texture analysis and LVQ neural network

Liu, Jianli; Zuo, Baoqi; Zeng, Xiaoyang; Vroman, Philippe; Rabenasolo, Besoa

doi:10.1016/j.eswa.2009.07.049

Cited by 37 publications

(18 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The two-dimensional (2D) wavelet transform could be perceived as one-dimensional wavelet transform sequentially applied along the horizontal and vertical axes. 29,30 With the pyramid-structured wavelet transform, the original image first passes through the low-pass and high-pass decomposition filters to generate four lower resolution components. 19 One is called the low-low (LL1) smooth subimage, which is the approximation of the original image.…”

Section: Wavelet Transformmentioning

confidence: 99%

Seam detection of inhomogeneously textured fabrics based on wavelet transform

Zhu

Liu

Pan

et al. 2014

Textile Research Journal

View full text Add to dashboard Cite

A large amount of seam detection for inhomogeneously textured fabrics makes those performing it fatigued, which leads to misjudgments by human vision, especially for the seam detection of patterned inhomogeneously textured fabrics. The traditional wavelet texture analysis is no longer applicable to fabrics with inhomogeneous textures and irregular patterns. In this paper, a novel mean weighting factor is proposed to obtain an optimized discriminant measure to detect the fabric seams. Firstly, the wavelet coefficients are extracted in individual decomposition levels. Then a mean weighting factor is calculated on the use of the difference of the coefficient values between two consecutive decomposition levels, and a better discriminant measure for seam detection is obtained. Lastly, a thresholding process is used to segment the seam information from the background. The experimental results show that the proposed approach effectively carries out the seam detection in the inhomogeneously textured fabric images.Keywords inhomogeneously textured fabrics, seam detection, wavelet transform, mean weighting factor, discriminant measureTo improve the quality of velvet fabrics, the cutting process is a necessary step to make the surface uniform. Figure 1 shows the cutting process of velvet fabrics. The thickness of the seam is over 3.5 mm, which is slightly larger than the distance between the driving roller and blade (about 2-3 mm). The blade at a high running speed (the speed is more than 1000 rpm) will be damaged by the fabric seams if they contact with each other. Then fabric defects will occur when the following fabrics are operated by the damaged blade. Besides, the blade in the cutting machine is fragile and expensive. The seam will not only lead to fabric disqualification, but also causes equipment loss. So the detection of velvet fabric seams plays a key role in the cutting process.Traditionally, manual inspectors control the blade going up and down to avoid fabric seams. This manual approach has drawbacks with regard to identifying seams in terms of accuracy, consistency and efficiency, as inspectors are subject to visual fatigue or brain delayed-response after a long period of working and thus seam misjudgments are often produced. To eliminate the human factors to increase product quality, an advanced system of setting a thickness sensor alongside the blade is used to detect the fabric seams in some factories. A set of control system makes the blade rise up when the seam goes through the thickness sensor. However, with the increase of velvet fabric thickness, this system has faced limitations on the induction of the change of the thickness. There are urgent industrial demands to investigate the machine vision technique 1,2 for seam detection.In fact, machine vision has been widely used for the fabric defect detection. 3-7 Various algorithms designed for fabric defect detection [8][9][10][11][12][13] have been developed to tackle the detection problem, such as the Fourier transform, 14,15 the Gabor filters 16,17 and...

show abstract

Section: Wavelet Transformmentioning

confidence: 99%

Seam detection of inhomogeneously textured fabrics based on wavelet transform

Zhu

Liu

Pan

et al. 2014

Textile Research Journal

View full text Add to dashboard Cite

show abstract

“…The input layer contains one node for each input feature, the output layer contains one node for each class. The process methodology can be found in [17], [18].…”

Section: Learning Vector Quantization Networkmentioning

confidence: 99%

Smart grids power quality analysis based in classification techniques and higher-order statistics: Proposal for photovoltaic systems

Palomares-Salas

Rosa

Agüera-Pérez

et al. 2015

2015 IEEE International Conference on Industrial Technology (ICIT)

View full text Add to dashboard Cite

This paper compares classification techniques to identify several power quality disturbances in a frame of smart metering design for smart grids with high penetration of PV systems. These techniques are: Linear Discriminant Analysis (LDA), Nearest Neighbor Method (kNN), Learning Vector Quantization (LVQ) and Support Vector Machine (SVM). For this purpose, fourteen power-quality features based in higher-order statistics are used to assist classification. Special attention is paid to the spectral kurtosis, whose nature enables measurement options related to the impulsiveness of the power quality events. The best technique of those compared is selected according to correlation and mistake rates. Results clearly reveal the potential capability of the methodology in classifying the single disturbances. Concretely, the SVM classifier obtained an average correlation rate of 99%. Hence, concluding that it is a robust classification method.978-1-4799-7800-7/15/$31.00 ©2015 IEEE

show abstract

“…The visual quality inspection of nonwovens based on digital image processing has been an obvious trend with the rapid development of nonwoven industry [1], In this work, a novel system to recognize the visual quality of nonwovens that involves wavelet texture analysis and K-NN classifier is presented. Each nonwoven image is decomposed into 4 levels with wavelet base db8 and coif4, and two textural features, norm-1 L' and norm-2 L calculated from the wavelet coefficients of each high frequency subband, are used to train and test K-NN classifier.…”

Section: Introductionmentioning

confidence: 99%

Wavelet Texture Analysis and K-NN Classifier for Nonwovens' Visual Quality Recognition

Liu¹,

Zuo²,

Zeng³

et al. 2010

International Journal of Nonlinear Sciences and Numerical Simulation

Self Cite

View full text Add to dashboard Cite

A novel system for the visual quality recognition of nonwovens is presented by combining wavelet texture analysis and K-nearest neighbor (K-NN) classifier and its numerical simulation is also illustrated. In this research, 625 nonwovens images of 5 different grades, 125 samples of each grade, are decomposed into 4 levels with wavelet base db8 and coif4, and then two energy signatures, norm-1 L' and norm-2 L 2 are calculated from wavelet coefficients of each high frequency subband to train and test K-NN classifier. When the nonwoven images are decomposed into 3 levels and 9 L ] s are extracted as textural features, the highest recognition accuracy of numerical simulation with K-NN classifier is as high as 96%.

show abstract

Nonwoven uniformity identification using wavelet texture analysis and LVQ neural network

Cited by 37 publications

References 6 publications

Seam detection of inhomogeneously textured fabrics based on wavelet transform

Seam detection of inhomogeneously textured fabrics based on wavelet transform

Smart grids power quality analysis based in classification techniques and higher-order statistics: Proposal for photovoltaic systems

Wavelet Texture Analysis and K-NN Classifier for Nonwovens' Visual Quality Recognition

Contact Info

Product

Resources

About