In ultra-high molecular weight polyethylene (UHMW-PE) fiber production, surface defects such as stiffness, hairiness, and broken filaments are common, and these defects caused by production process or equipment problems seriously affect product quality. Traditional manual visual inspection methods are inefficient and limited in accuracy, which do not meet the demand of high-speed production. In this study, an online quality inspection algorithm based on machine vision is proposed, in which a high-performance industrial camera captures the original polyethylene fiber images, and the fiber defects are effectively identified after image preprocessing, fiber segmentation, localization, feature extraction, and anomaly identification. Experimental validation shows that the algorithm is efficient, accurate, and reliable, significantly improving detection efficiency and accuracy and providing an effective means for UHMW-PE fiber quality control.