郑州大学学报（理学版）2025，Vol.57Issue(3)：1-11,11.DOI:10.13705/j.issn.1671-6841.2023239
基于少样本学习的表面缺陷检测方法综述
Review of Surface Defect Detection Methods Based on Few-shot Learning
摘要
Abstract
In some industrial scenarios,insufficient defect samples and labeling time-consuming and la-bor-intensive defects,limit the application of machine vision methods in surface defect detection.Tech-nologies of industrial defect detection based on few-shot learning were introduced from three aspects:im-age acquisition,image processing,and defect detection.Firstly,defect detection methods were divided into traditional surface defect detection methods and few-shot deep learning based defect detection meth-ods.The traditional surface defect detection method was based on the manually extracted features to iden-tify defects,which could be divided into three parts:defect segmentation,artificial feature extraction and defect recognition.Few-shot deep learning based industrial defect detection methods include data enhancement,transfer learning,model fine-tuning,semi-supervised learning,weakly supervised learn-ing,unsupervised learning methods,etc.Secondly,some commonly used defect detection datasets and evaluation criteria of detection results were introduced.Finally,the existing problems and future research directions of few-shot learning based surface defect detection were discussed.关键词
缺陷检测/少样本学习/机器视觉/深度学习Key words
defect detection/few-shot learning/machine vision/deep learning分类
计算机与自动化引用本文复制引用
陈丽,殷湘婷,靳启帆,姜晓恒,酒明远,徐明亮..基于少样本学习的表面缺陷检测方法综述[J].郑州大学学报（理学版）,2025,57(3):1-11,11.基金项目
国家自然科学基金项目(62202433,U21B2037,62272422,62172371,U22B2051) （62202433,U21B2037,62272422,62172371,U22B2051）
国家重点研发计划项目(YFB3301504) （YFB3301504）
河南省博士后基金项目(202103111) （202103111）
河南省自然科学基金项目(22100002) （22100002）
