Fabric Defect Detection Using Fuzzy Inductive Reasoning Based on Image Histogram Statistic Variables

This paper deals with the fuzzy inductive reasoning (FIR) for fabric defect detection. Based on linear and regular texture of the fabric, we first extract histogram statistic variables as the distinguishing features between faultless and faulty fabric images. By applying FIR to the histogram statistic variables and subtracting the class values of the real statistic variables and the predicted class values using the qualitative model, cumulative errors are computed, which are used to determine if a defect has been detected. Simulation experiments show that the proposed method can achieve a robust and accurate detection of fabric defects.