A ResNet50-Based Method for Classifying Surface Defects in Hot-Rolled Strip Steel
Hot-rolled strip steel is widely used in automotive manufacturing, chemical and home appliance industries, and its surface quality has a great impact on the quality of the final product. In the manufacturing process of strip steel, due to the rolling process and many other reasons, the surface of ho...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2021-09-01
|
| Series: | Mathematics |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2227-7390/9/19/2359 |
| _version_ | 1797515986155339776 |
|---|---|
| author | Xinglong Feng Xianwen Gao Ling Luo |
| author_facet | Xinglong Feng Xianwen Gao Ling Luo |
| author_sort | Xinglong Feng |
| collection | DOAJ |
| description | Hot-rolled strip steel is widely used in automotive manufacturing, chemical and home appliance industries, and its surface quality has a great impact on the quality of the final product. In the manufacturing process of strip steel, due to the rolling process and many other reasons, the surface of hot rolled strip steel will inevitably produce slag, scratches and other surface defects. These defects not only affect the quality of the product, but may even lead to broken strips in the subsequent process, seriously affecting the continuation of production. Therefore, it is important to study the surface defects of strip steel and identify the types of defects in strip steel. In this paper, a scheme based on ResNet50 with the addition of FcaNet and Convolutional Block Attention Module (CBAM) is proposed for strip defect classification and validated on the X-SDD strip defect dataset. Our solution achieves a classification accuracy of 94.11%, higher than more than a dozen other compared deep learning models. Moreover, to adress the problem of low accuracy of the algorithm in classifying individual defects, we use ensemble learning to optimize. By integrating the original solution with VGG16 and SqueezeNet, the recognition rate of oxide scale of plate system defects improved by 21.05 percentage points, and the overall defect classification accuracy improved to 94.85%. |
| first_indexed | 2024-03-10T06:55:55Z |
| format | Article |
| id | doaj.art-52834a34f9364d4abe0e35aaaa9e3a00 |
| institution | Directory Open Access Journal |
| issn | 2227-7390 |
| language | English |
| last_indexed | 2024-03-10T06:55:55Z |
| publishDate | 2021-09-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Mathematics |
| spelling | doaj.art-52834a34f9364d4abe0e35aaaa9e3a002023-11-22T16:29:11ZengMDPI AGMathematics2227-73902021-09-01919235910.3390/math9192359A ResNet50-Based Method for Classifying Surface Defects in Hot-Rolled Strip SteelXinglong Feng0Xianwen Gao1Ling Luo2College of Information Science and Engineering, Northeastern University, Shenyang 110819, ChinaCollege of Information Science and Engineering, Northeastern University, Shenyang 110819, ChinaMoviebook Technology Co., Ltd., Beijing 100027, ChinaHot-rolled strip steel is widely used in automotive manufacturing, chemical and home appliance industries, and its surface quality has a great impact on the quality of the final product. In the manufacturing process of strip steel, due to the rolling process and many other reasons, the surface of hot rolled strip steel will inevitably produce slag, scratches and other surface defects. These defects not only affect the quality of the product, but may even lead to broken strips in the subsequent process, seriously affecting the continuation of production. Therefore, it is important to study the surface defects of strip steel and identify the types of defects in strip steel. In this paper, a scheme based on ResNet50 with the addition of FcaNet and Convolutional Block Attention Module (CBAM) is proposed for strip defect classification and validated on the X-SDD strip defect dataset. Our solution achieves a classification accuracy of 94.11%, higher than more than a dozen other compared deep learning models. Moreover, to adress the problem of low accuracy of the algorithm in classifying individual defects, we use ensemble learning to optimize. By integrating the original solution with VGG16 and SqueezeNet, the recognition rate of oxide scale of plate system defects improved by 21.05 percentage points, and the overall defect classification accuracy improved to 94.85%.https://www.mdpi.com/2227-7390/9/19/2359hot rolled strip steeldeep learningsurface defectsdefect classification |
| spellingShingle | Xinglong Feng Xianwen Gao Ling Luo A ResNet50-Based Method for Classifying Surface Defects in Hot-Rolled Strip Steel Mathematics hot rolled strip steel deep learning surface defects defect classification |
| title | A ResNet50-Based Method for Classifying Surface Defects in Hot-Rolled Strip Steel |
| title_full | A ResNet50-Based Method for Classifying Surface Defects in Hot-Rolled Strip Steel |
| title_fullStr | A ResNet50-Based Method for Classifying Surface Defects in Hot-Rolled Strip Steel |
| title_full_unstemmed | A ResNet50-Based Method for Classifying Surface Defects in Hot-Rolled Strip Steel |
| title_short | A ResNet50-Based Method for Classifying Surface Defects in Hot-Rolled Strip Steel |
| title_sort | resnet50 based method for classifying surface defects in hot rolled strip steel |
| topic | hot rolled strip steel deep learning surface defects defect classification |
| url | https://www.mdpi.com/2227-7390/9/19/2359 |
| work_keys_str_mv | AT xinglongfeng aresnet50basedmethodforclassifyingsurfacedefectsinhotrolledstripsteel AT xianwengao aresnet50basedmethodforclassifyingsurfacedefectsinhotrolledstripsteel AT lingluo aresnet50basedmethodforclassifyingsurfacedefectsinhotrolledstripsteel AT xinglongfeng resnet50basedmethodforclassifyingsurfacedefectsinhotrolledstripsteel AT xianwengao resnet50basedmethodforclassifyingsurfacedefectsinhotrolledstripsteel AT lingluo resnet50basedmethodforclassifyingsurfacedefectsinhotrolledstripsteel |