基于改进U-Transformer模型的金刚石刀刃异常检测算法*

doi:10.16731/j.cnki.1671-3133.2025.07.015

现代制造工程 ›› 2025, Vol. 538 ›› Issue (7): 120-128.doi: 10.16731/j.cnki.1671-3133.2025.07.015

• 仪器仪表/检测/监控 • 上一篇下一篇

基于改进U-Transformer模型的金刚石刀刃异常检测算法^*

王大伟, 李丛, 朱长水

南京理工大学泰州科技学院计算机科学与工程学院,泰州 225300

收稿日期:2024-09-29 出版日期:2025-07-18 发布日期:2025-08-04
作者简介:王大伟,硕士,高级工程师,主要研究方向为图像处理、机器视觉。李丛,硕士,副教授,主要研究方向为智能信息处理。朱长水,硕士,副教授,主要研究方向为图像处理。E-mail:583705373@qq.com
基金资助:
^*国家自然科学基金资助项目(61272210);泰州市科技支撑计划(社会发展)项目(SSF20230056)

Anomaly detection algorithm for diamond blade based on improved U-Transformer model

WANG Dawei, LI Cong, ZHU Changshui

School of Computer Science and Engineering,Taizhou Institute of Science and Technology, Nanjing University of Science and Technology,Taizhou 225300,China

Received:2024-09-29 Online:2025-07-18 Published:2025-08-04

摘要/Abstract

摘要： 针对金刚石刀刃缺陷特征差异大、缺陷样本少的问题,提出了一种基于改进U-Transformer特征重建模型的金刚石刀刃异常检测算法。该方法仅需使用正常样本训练即可完成异常区域的检测与定位。首先,利用冻结的预训练深度卷积神经网络(Convolutional Neural Networks,CNN)模型提取多尺度融合特征,放大正常图像与异常图像的差异;然后,构建基于U型Transformer结构的编码器-解码器特征重建模型,计算重建特征与输入特征的特征相似性,生成相似性响应图;最后,为消除正常区域的噪声响应,利用多层感知机(Multi-Layer Perceptron,MLP)网络估计异常比例因子,修正相似性响应图,得到异常分数图。实验结果表明,提出的方法在金刚石刀刃缺陷数据集上Image AUROC指标为0.989,Piexl AUROC指标为0.992,能够满足金刚石刀刃异常检测需求。

关键词: 金刚石刀刃, 异常检测, U-Transformer模型, 预训练, 多层感知机网络, 特征重建

Abstract: Aiming at the problems of large differences in diamond blade defect features and limited defect samples,an anomaly detection algorithm for diamond blade based on the improved U-Transformer feature reconstruction model was proposed.This method only needs to be trained with normal samples to complete the detection and localization of abnormal areas. Firstly,a frozen pre-trained deep Convolutional Neural Networks (CNN) model was used to extract multi-scale fusion features,amplifying the differences between normal and abnormal images. Then,the encoder-decoder feature reconstruction model based on the U-Transformer structure was constructed,and the feature similarity between the reconstructed features and the input features was calculated to generate the similarity response map. Finally,to eliminate the noise responses in the normal region,the similarity response map was corrected to obtain the anomaly score map by estimating the anomaly scale factor using the Multi-Layer Perceptron (MLP) network. The experimental results demonstrate that the proposed method achieves an Image AUROC index of 0.989 and a Pixel AUROC index of 0.992 on the diamond blade defect dataset,meeting the requirements for diamond blade anomaly detection.

Key words: diamond blade, anomaly detection, U-Transformer model, pre-training, MLP network, feature reconstruction

中图分类号:

TP391.41

王大伟, 李丛, 朱长水. 基于改进U-Transformer模型的金刚石刀刃异常检测算法^*[J]. 现代制造工程, 2025, 538(7): 120-128.

WANG Dawei, LI Cong, ZHU Changshui. Anomaly detection algorithm for diamond blade based on improved U-Transformer model[J]. Modern Manufacturing Engineering, 2025, 538(7): 120-128.

参考文献

[1] 白清顺,张亚博,王永旭,等.微裂纹缺陷对CVD金刚石涂层微刀具损伤失效的影响研究[J]. 表面技术,2021,50(2):355-362.
[2] 张杨,高兴宇,党艳阳,等.基于三维图像处理的车刀磨损缺陷检测方法研究[J]. 机床与液压,2023,51(24):43-47[2024-07-10].
[3] 何钢,朱峰,邹华涛,等.基于改进Frangi滤波的芯片表面缺陷检测算法[J/OL]. 计算机集成制造系统:1-14[2024-07-10]. https://doi.org/10.13196/j.cims.2023.0397.
[4] 宿磊,王立建,祁阳,等.基于IADSA深度迁移网络的金属表面缺陷检测[J]. 机械工程学报,2023(24):59.
[5] SU B,CHEN H Y,CHEN P,et al.Deep Learning-Based Solar-Cell Manufacturing Defect Detection With Complementary Attention Network[J]. IEEE Transactions on Industrial Informatics,2020(99):1.
[6] ZHOU Y,JI A,ZHANG L.Sewer defect detection from 3D point clouds using a transformer-based deep learning model[J]. Automation in Construction,2022(4):136.
[7] TANG W,YANG Q,XIONG K,et al.Deep learning based automatic defect identification of photovoltaic module using electroluminescence images[J]. Solar Energy,2020,201:453-460.
[8] DAVARI N,AKBARIZADEH G,MASHHOUR E,et al.Corona Detection and Power Equipment Classification Based on GoogleNet-AlexNet:An Accurate and Intelligent Defect Detection Model Based on Deep Learning for Power Distribution Lines[J]. IEEE Transactions on Power Delivery,2022,37(8):2766-2774.
[9] XUE W, ZHAO C, FU W, et al.Micro Vision-based Sharpening Quality Detection of Diamond Tools[J]. Intelligent Robotics and Applications,2022,2835(13458):245-252.
[10] ZHAO C,XUE W,FU W,et al.Defect Sample Image Generation Method Based on GANs in Diamond Tool Defect Detection[J]. IEEE Transactions on Instrumentation and Measurement,2023(72):1-9.
[11] 刘浩,陈再良,王善翔.基于改进YOLOv3算法的刀具表面缺陷检测[J]. 组合机床与自动化加工技术,2021(11):87-90.
[12] 姜金涛,丁坤,王志花,等.基于改进嵌入向量相似性的织物异常检测方法[J]. 毛纺科技,2023,51(6):73-80.
[13] TSAI D M,JEN P H.Autoencoder-based anomaly detection for surface defect inspection[J]. Advanced Engineering Informatics,2021(48):101272.
[14] CHOW J K,SU Z,WU J,et al.Anomaly detection of defects on concrete structures with the convolutional autoencoder[J]. Advanced Engineering Informatics,2020,45(3):101105.
[15] KIMURA M,YANAGIHARA T.Semi-supervised Anomaly Detection Using GANs for Visual Inspection in Noisy Training Data[C]//Computer Vision-ACCV 2018 Workshops:14th Asian Conference on Computer Vision.[S.l.] :Springer International Publishing,2019:373-385.
[16] RIBEIROM,LAZZARETTI A,LOPES H.A study of deep convolutional auto-encoders for anomaly detection in videos[J]. Pattern Recognition Letters,2017,105:13-22.
[17] CHAABI M,HAMLICH M,GAROUANI M.Product defect detection based on convolutional autoencoder and one-class classification[J]. IAES International Journal of Artificial Intelligence,2023,12(2):912-920.
[18] SALEHI M,SADJADI N,BASELIZADEH S,et al.Multiresolution Knowledge Distillation for Anomaly Detection[C]//Computer Vision and Pattern Recognition.[S.l.] :IEEE,2021.
[19] DEFARD T,SETKOV A,LPESCH A,et al.PaDiM:A Patch Distribution Modeling Framework for Anomaly Detection and Localization[C]//International Conference on Pattern Recognition.Cham:Springer International Publishing,2021:475-489.
[20] HAN K,WANG Y,CHEN H,et al.A Survey on Vision Transformer[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence,2023,45:87-110.
[21] CARION N,MASSA F,SYNNAEVE G,et al.End-to-End Object Detection with Transformers[C]//European conference on computer vision.Cham:Springer International Publishing,2020:213-229.
[22] YANG F,YANG H,FU J,et al.Learning Texture Transformer Network for Image Super-Resolution[C]//CVF Conference on Computer Vision and Pattern Recognition (CVPR).Piscataway:IEEE,2020:5791-5800.
[23] CHEN L,YOU Z,ZHANG N,et al.UTRAD:Anomaly detection and localization with U-Transformer[J]. Neural Networks,2022,147:53-62.
[24] 张玥,陈锡伟,陈梦丹,等.基于对比学习生成对抗网络的无监督工业品表面异常检测[J]. 电子测量与仪器学报,2023,37(10):193-201.
[25] GUDOVSKIY D,ISHIZAKA S,KOZUKA K.CFLOW-AD:Real-Time Unsupervised Anomaly Detection with Localization via Conditional Normalizing Flows[C]//CVF Winter confer-ence on Applications of Computer Vision (WACV).Piscata-way:IEEE,2022:98-107.
[26] ROTH K,PEMULA L,ZEPEDA J,et al.Towards Total Recall in Industrial Anomaly Detection[C]//CVF Conference on Computer Vision and Pattern Recognition (CVPR).Piscataway:IEEE,2022:14318-14328.

基于改进U-Transformer模型的金刚石刀刃异常检测算法^*

Anomaly detection algorithm for diamond blade based on improved U-Transformer model

PDF

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 1

编辑推荐

Metrics