A note on neurofractal-based defect recognition and classification in nonwoven web images

摘要

This paper introduces the off-line neurofractal method developed for defect detection and classification in thermal-bond nonwoven web images using box counting dimension as feature extractor and backpropagation neural network algorithm as defect classifier. The results of applying the proposed methodology on nonwoven web images show that defects are recognized and classified with high accuracy.