Abstract :
There is not a prefect method for defect feature testing of fabrics using computer at present. In this paper, a 3-D gray feature change model is proposed after a great deal of experiments, and fabric feature is tested and measured. All these provide the theory and data basis for the research of automatic defect recognition. Firstly, mathematic description of 3-D gray feature change model is given, and detail testing method of fabric feature is introduced. Then, different experiments are designed according to the energy, variance, entropy, and range, and expression effect is tested and 3-D gray feature change graphs are drawn. Finally, according to these testing results, detail applicable defect category with these features is given, and result is concluded that 3-D gray feature change model can test fabric defect feature better.
Keywords :
computer testing; entropy; fabrics; feature extraction; inspection; production engineering computing; quality control; 3-D gray feature change model; automatic defect recognition; computer testing method; defect features; entropy; fabric feature; features extraction; mathematic description; quality inspection; textile industry production; Clustering algorithms; Concrete; Entropy; Fabrics; Feature extraction; Inspection; Mathematical model; Mathematics; Pixel; Testing; Defect; Fabric; Feature; Testing;