AUTOMATED INSPECTION OF TEXTILE FABRICS USING TEXTURAL MODELS

This correspondence is concerned with the problem of textile fabric inspection using the visual textural properties of the fabric. The problem here is to detect and locate the various kinds of defects that might be present in a given fabric sample based on an image of the fabric. Stochastic models are used to model the visual fabric texture. We use the Gaussian Markov Random Field (GMRF) to model the texture image of a nondefective fabric texture. The GMRF model is fully described by a compact set of parameters, and is a synthesis model. The inspection problem is cast as a statistical hypothesis testing problem on statistics derived from the model. The image of the fabric patch to be inspected is partitioned into nonoverlapping windows of size N x N, where each window is classified as defective or nondefective based on a likelihood ratio test of size alpha. This test requires the computation of the maximum likelihood estimates (MLE) of the model parameters in each window. This is somewhat computationally intensive and time consuming. To overcome this problem, the test is recast in terms of the sufficient statistics associated with the model parameters. The sufficient statistics is easily computable for any sample. Finally, we generalize the test when the model parameters of the fabric can not be obtained beforehand, and hence are assumed unknown.