Recent advances in fabric defect detection algorithms

https://elibrary.ru/FQBCXE

Abstract

Against the backdrop of the textile industry’s transition to smart manufacturing, fabric defect detection has become a critical step in ensuring product quality, enhancing production efficiency and reducing operational costs. Traditional manual inspection relies primarily on visual judgement; whilst it allows for the immediate identification of obvious defects, it suffers from low efficiency, a high degree of subjectivity and susceptibility to visual fatigue among inspectors, resulting in a high rate of missed detection of minor defects. Consequently, the development of automated inspection technology has become an inevitable trend for improving fabric quality and reducing labour costs. With the rapid advancement of computer vision and machine learning technologies, particularly the increasing maturity of deep learning methods, intelligent algorithm-based fabric defect detection has gradually become a core tool in modern textile quality control. This paper systematically reviews research progress in the field of fabric defect detection, categorising existing methods into two main groups: those based on traditional image processing and those based on deep learning. On this basis, a comparative analysis of the two categories is conducted, exploring in depth their respective technical advantages and limitations, thereby providing a reference for future research and industrial applications.