2-D Iteratively Reweighted Least Squares Lattice Algorithm and Its Application to Defect Detection in Textured Images

Abstract

-In this paper, a 2-D iteratively reweighted least squares lattice algorithm, which is robust to the outliers, is introduced and is applied to defect detection problem in textured images. First, the philosophy of using different optimization functions that results in weighted least squares solution in the theory of 1-D robust regression is extended to 2-D. Then a new algorithm is derived which combines 2-D robust regression concepts with the 2-D recursive least squares lattice algorithm. With this approach, whatever the probability distribution of the prediction error may be, small weights are assigned to the outliers so that the least squares algorithm will be less sensitive to the outliers. Implementation of the proposed iteratively reweighted least squares lattice algorithm to the problem of defect detection in textured images is then considered. The performance evaluation, in terms of defect detection rate, demonstrates the importance of the proposed algorithm in reducing the effect of the outliers that generally correspond to false alarms in classification of textures as defective or nondefective.

A Method of Fabric Defect Detection Using Local Contrast Deviation

“…Image segmentation in image-based inspection systems is the key to automatic detection of fabric defects. The texture and shape features of fabric defects can be extracted in the spatial domain using methods such as gray concurrence matrix, Markov random fields and mathematical morphology, ICA, and textural models, and then they can be classified using preset thresholds or artificial neural networks [5][6][7][8][9]. Fabric images can also be transformed to other domains using the fast Fourier transform, the Gabor transform, or the wavelet transform for locating defects in the images [10][11][12][13].…”

Model based defect detection problem: Particle filter approach

“…Here the main concern is to detect defects in fabrics during the roll of the web product and to warn the manufacturer before the defective product reaches the market. Considering that human vision inspection has a performance of 80 per cent, at best, designing computer related visual inspection systems has became a requirement for textile fabric production plants [1][2][3].…”

“…Model based texture analysis is commonly addressed by modeling texture under analysis in terms of Markov random fields [4,5] or autoregressive (AR) and/or autoregressive moving average (ARMA) fields [2,6]. In AR/ARMA based modeling, linear prediction filters such as lattice filters [2,7] or Kalman filters [8] can be used to estimate the model parameters that represent the complex texture in terms of a small number of parameters. In this paper, we bring a new solution to the model based texture analysis and hence to the defect detection problem.…”