Result Weighting-Based Resnet Feature Pyramid Network Architecture for Surface Defect Detection
Surface defect detection is very important in manufacturing systems to ensure high quality products. Unlike manual inspections under human supervision, automatic surface defect detection is both efficient and highly accurate. In this study, a Result Weighting-based Resnet Feature Pyramid Network (SA...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Gazi University
2021-12-01
|
| Series: | Gazi Üniversitesi Fen Bilimleri Dergisi |
| Subjects: | |
| Online Access: | https://dergipark.org.tr/tr/download/article-file/2074414 |
| _version_ | 1828026060996870144 |
|---|---|
| author | Hüseyin ÜZEN Muammer TÜRKOĞLU Davut HANBAY |
| author_facet | Hüseyin ÜZEN Muammer TÜRKOĞLU Davut HANBAY |
| author_sort | Hüseyin ÜZEN |
| collection | DOAJ |
| description | Surface defect detection is very important in manufacturing systems to ensure high quality products. Unlike manual inspections under human supervision, automatic surface defect detection is both efficient and highly accurate. In this study, a Result Weighting-based Resnet Feature Pyramid Network (SA-RÖPA) model has been developed for automatic pixel-level surface defect detection. In the first stage of the proposed model, the pre-trained Resnet50 network was used, and feature maps were extracted from the different levels of this network. In the second stage, Feature Pyramid Model was applied to these feature maps in order to hierarchically share important information in defect detection. In the third stage, 4 different error detection results were obtained by using these feature maps. In the last stage, four different results obtained using the developed Result Weighting (SA) module were effectively combined. The proposed SA-ROPA model has been tested with MT, MVTex-Doku, and AITEX datasets, which are widely used in defect detection studies. In experimental studies, the mIoU value obtained for the MT and AITEX datasets using the proposed model was calculated as 79.92%, 76.37%, and 82.72%, respectively. These results have shown that the proposed SA- ROPA model is more successful than other state-of-the-art models. |
| first_indexed | 2024-04-10T13:13:33Z |
| format | Article |
| id | doaj.art-38e79b1cc1e34690b91e8d7faea5e6af |
| institution | Directory Open Access Journal |
| issn | 2147-9526 |
| language | English |
| last_indexed | 2024-04-10T13:13:33Z |
| publishDate | 2021-12-01 |
| publisher | Gazi University |
| record_format | Article |
| series | Gazi Üniversitesi Fen Bilimleri Dergisi |
| spelling | doaj.art-38e79b1cc1e34690b91e8d7faea5e6af2023-02-15T16:12:30ZengGazi UniversityGazi Üniversitesi Fen Bilimleri Dergisi2147-95262021-12-019476077210.29109/gujsc.1021785Result Weighting-Based Resnet Feature Pyramid Network Architecture for Surface Defect DetectionHüseyin ÜZENhttps://orcid.org/0000-0002-0998-2130Muammer TÜRKOĞLUhttps://orcid.org/0000-0002-2377-4979Davut HANBAYhttps://orcid.org/0000-0003-2271-7865Surface defect detection is very important in manufacturing systems to ensure high quality products. Unlike manual inspections under human supervision, automatic surface defect detection is both efficient and highly accurate. In this study, a Result Weighting-based Resnet Feature Pyramid Network (SA-RÖPA) model has been developed for automatic pixel-level surface defect detection. In the first stage of the proposed model, the pre-trained Resnet50 network was used, and feature maps were extracted from the different levels of this network. In the second stage, Feature Pyramid Model was applied to these feature maps in order to hierarchically share important information in defect detection. In the third stage, 4 different error detection results were obtained by using these feature maps. In the last stage, four different results obtained using the developed Result Weighting (SA) module were effectively combined. The proposed SA-ROPA model has been tested with MT, MVTex-Doku, and AITEX datasets, which are widely used in defect detection studies. In experimental studies, the mIoU value obtained for the MT and AITEX datasets using the proposed model was calculated as 79.92%, 76.37%, and 82.72%, respectively. These results have shown that the proposed SA- ROPA model is more successful than other state-of-the-art models.https://dergipark.org.tr/tr/download/article-file/2074414surface defect detectionpyramid feature networkconvolutional neural networksegmentation |
| spellingShingle | Hüseyin ÜZEN Muammer TÜRKOĞLU Davut HANBAY Result Weighting-Based Resnet Feature Pyramid Network Architecture for Surface Defect Detection Gazi Üniversitesi Fen Bilimleri Dergisi surface defect detection pyramid feature network convolutional neural network segmentation |
| title | Result Weighting-Based Resnet Feature Pyramid Network Architecture for Surface Defect Detection |
| title_full | Result Weighting-Based Resnet Feature Pyramid Network Architecture for Surface Defect Detection |
| title_fullStr | Result Weighting-Based Resnet Feature Pyramid Network Architecture for Surface Defect Detection |
| title_full_unstemmed | Result Weighting-Based Resnet Feature Pyramid Network Architecture for Surface Defect Detection |
| title_short | Result Weighting-Based Resnet Feature Pyramid Network Architecture for Surface Defect Detection |
| title_sort | result weighting based resnet feature pyramid network architecture for surface defect detection |
| topic | surface defect detection pyramid feature network convolutional neural network segmentation |
| url | https://dergipark.org.tr/tr/download/article-file/2074414 |
| work_keys_str_mv | AT huseyinuzen resultweightingbasedresnetfeaturepyramidnetworkarchitectureforsurfacedefectdetection AT muammerturkoglu resultweightingbasedresnetfeaturepyramidnetworkarchitectureforsurfacedefectdetection AT davuthanbay resultweightingbasedresnetfeaturepyramidnetworkarchitectureforsurfacedefectdetection |