Fabric defect detection algorithm based on PHOG and SVM

Abstract

In order to effectively improve the detection probability for different types of fabrics and defects, a fabric defect detection method based on pyramid histogram of edge orientation gradients (PHOG) and support vector machine (SVM) has been proposed. The algorithm combines fabric texture statistical method and machine learning method. It has two main parts, namely the feature extraction and classification. The detection process mainly includes image segmentation, PHOG feature extraction, SVM model training and detection classification. The simulation results show that, based on the detection rate and the false alarm rate, the algorithm has a good detection and classification effect, has a certain robustness, and can be applied to the actual production department.