Improving Industrial Quality Control: A Transfer Learning Approach to Surface Defect Detection
To automate the quality control of painted surfaces of heating devices, an automatic defect detection and classification system was developed by combining deflectometry and bright light-based illumination on the image acquisition, deep learning models for the classification of non-defective (OK) and...
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2025-01-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/25/2/527 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832587502792212480 |
|---|---|
| author | Ângela Semitela Miguel Pereira António Completo Nuno Lau José P. Santos |
| author_facet | Ângela Semitela Miguel Pereira António Completo Nuno Lau José P. Santos |
| author_sort | Ângela Semitela |
| collection | DOAJ |
| description | To automate the quality control of painted surfaces of heating devices, an automatic defect detection and classification system was developed by combining deflectometry and bright light-based illumination on the image acquisition, deep learning models for the classification of non-defective (OK) and defective (NOK) surfaces that fused dual-modal information at the decision level, and an online network for information dispatching and visualization. Three decision-making algorithms were tested for implementation: a new model built and trained from scratch and transfer learning of pre-trained networks (ResNet-50 and Inception V3). The results revealed that the two illumination modes employed widened the type of defects that could be identified with this system, while maintaining its lower computational complexity by performing multi-modal fusion at the decision level. Furthermore, the pre-trained networks achieved higher accuracies on defect classification compared to the self-built network, with ResNet-50 displaying higher accuracy. The inspection system consistently obtained fast and accurate surface classifications because it imposed OK classification on models trained with images from both illumination modes. The obtained surface information was then successfully sent to a server to be forwarded to a graphical user interface for visualization. The developed system showed considerable robustness, demonstrating its potential as an efficient tool for industrial quality control. |
| format | Article |
| id | doaj-art-76b3c35cc6924ecb9e1ebbd72b0b1da6 |
| institution | Kabale University |
| issn | 1424-8220 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj-art-76b3c35cc6924ecb9e1ebbd72b0b1da62025-01-24T13:49:14ZengMDPI AGSensors1424-82202025-01-0125252710.3390/s25020527Improving Industrial Quality Control: A Transfer Learning Approach to Surface Defect DetectionÂngela Semitela0Miguel Pereira1António Completo2Nuno Lau3José P. Santos4Centre of Mechanical Technology and Automation (TEMA), Department of Mechanical Engineering, University of Aveiro, 3810-193 Aveiro, PortugalCentre of Mechanical Technology and Automation (TEMA), Department of Mechanical Engineering, University of Aveiro, 3810-193 Aveiro, PortugalCentre of Mechanical Technology and Automation (TEMA), Department of Mechanical Engineering, University of Aveiro, 3810-193 Aveiro, PortugalIntelligent Systems Associate Laboratory (LASI), 4800-058 Guimarães, PortugalCentre of Mechanical Technology and Automation (TEMA), Department of Mechanical Engineering, University of Aveiro, 3810-193 Aveiro, PortugalTo automate the quality control of painted surfaces of heating devices, an automatic defect detection and classification system was developed by combining deflectometry and bright light-based illumination on the image acquisition, deep learning models for the classification of non-defective (OK) and defective (NOK) surfaces that fused dual-modal information at the decision level, and an online network for information dispatching and visualization. Three decision-making algorithms were tested for implementation: a new model built and trained from scratch and transfer learning of pre-trained networks (ResNet-50 and Inception V3). The results revealed that the two illumination modes employed widened the type of defects that could be identified with this system, while maintaining its lower computational complexity by performing multi-modal fusion at the decision level. Furthermore, the pre-trained networks achieved higher accuracies on defect classification compared to the self-built network, with ResNet-50 displaying higher accuracy. The inspection system consistently obtained fast and accurate surface classifications because it imposed OK classification on models trained with images from both illumination modes. The obtained surface information was then successfully sent to a server to be forwarded to a graphical user interface for visualization. The developed system showed considerable robustness, demonstrating its potential as an efficient tool for industrial quality control.https://www.mdpi.com/1424-8220/25/2/527automated quality controlilluminationtransfer learningdefect detection and classificationResNet-50 |
| spellingShingle | Ângela Semitela Miguel Pereira António Completo Nuno Lau José P. Santos Improving Industrial Quality Control: A Transfer Learning Approach to Surface Defect Detection Sensors automated quality control illumination transfer learning defect detection and classification ResNet-50 |
| title | Improving Industrial Quality Control: A Transfer Learning Approach to Surface Defect Detection |
| title_full | Improving Industrial Quality Control: A Transfer Learning Approach to Surface Defect Detection |
| title_fullStr | Improving Industrial Quality Control: A Transfer Learning Approach to Surface Defect Detection |
| title_full_unstemmed | Improving Industrial Quality Control: A Transfer Learning Approach to Surface Defect Detection |
| title_short | Improving Industrial Quality Control: A Transfer Learning Approach to Surface Defect Detection |
| title_sort | improving industrial quality control a transfer learning approach to surface defect detection |
| topic | automated quality control illumination transfer learning defect detection and classification ResNet-50 |
| url | https://www.mdpi.com/1424-8220/25/2/527 |
| work_keys_str_mv | AT angelasemitela improvingindustrialqualitycontrolatransferlearningapproachtosurfacedefectdetection AT miguelpereira improvingindustrialqualitycontrolatransferlearningapproachtosurfacedefectdetection AT antoniocompleto improvingindustrialqualitycontrolatransferlearningapproachtosurfacedefectdetection AT nunolau improvingindustrialqualitycontrolatransferlearningapproachtosurfacedefectdetection AT josepsantos improvingindustrialqualitycontrolatransferlearningapproachtosurfacedefectdetection |