福建电脑2024，Vol.40Issue(7)：37-40,4.DOI:10.16707/j.cnki.fjpc.2024.07.007
生成对抗网络在表面缺陷生成中的应用
Application of Generative Adversarial Network in Surface Defect Generation
摘要
Abstract
The object detection algorithm based on deep learning is widely used in the field of industrial product surface defect detection,but the constructed model requires a large amount of labeled product defect data.To reduce the cost of obtaining defect data,this paper proposes a surface defect generation algorithm based on generative adversarial networks.Through this algorithm,defect data that is closer to the true distribution can be generated.The experimental results show that the generated defect images are very realistic and contain a certain degree of discriminability.Using them as samples to participate in defect detection model training can produce regularization effects,thereby improving the robustness and generalization ability of defect detection.关键词
产品缺陷/表面缺陷检测/数据增强/缺陷生成算法Key words
Product Defects/Surface Defect Detection/Data Augmentation/Defect Generation Algorithm分类
信息技术与安全科学引用本文复制引用
刘日仙..生成对抗网络在表面缺陷生成中的应用[J].福建电脑,2024,40(7):37-40,4.基金项目
本文得到浙江省教育厅项目(No.Y201941616)资助. （No.Y201941616）
