Procedural synthetic training data generation for AI-based defect detection in industrial surface inspection

Enthalten in
Procedia CIRP 
Volume
107
Start Page
1101
End Page
1106
Citation
Procedia CIRP 107: 1101-1106 (2022)
Contribution to Conference
55th CIRP Conference on Manufacturing Systems, CIRP CMS 2022 
Publisher DOI
10.1016/j.procir.2022.05.115
Scopus ID
2-s2.0-85132247130
Publisher
Elsevier
Peer Reviewed
true
Supervised machine learning methods are increasingly used for detecting defects in automated visual inspection systems. However, these methods require large quantities of annotated image data of the surface being inspected, including images of defective surfaces. In industrial contexts, it is difficult to collect the latter since acquiring sufficient image data of defective surfaces is costly and time-consuming. Additionally, gathered datasets tend to contain selection-bias, e.g. under representation of certain defect classes, and therefore result in insufficient training data quality. Synthetic training data is a promising alternative as it can be easily generated unbiasedly and in large quantities. In this work, we present a procedural pipeline for generating training data based on physically based renderings of the object under inspection. Defects are being introduced as 3D-models on the surface of the object. The generator provides the ability to randomize object and camera parameters within given intervals, allowing the user to use the domain randomization technique to bridge the domain gap between the synthetic data and the real world. Experiments suggest that the data generated in this way can be beneficial to training defect detection models.
Schlagworte
Synthetic training data
machine learning
surface inspection
industrial quality control
domain randomization
MLE@TUHH
DDC Class
600: Technik
620: Ingenieurwissenschaften
Projekt(e)
Verfahren zur automatischen Qualitätskontrolle mittels robotergeführten Videoendoskopen 
Funding Organisations
Bundesministerium für Wirtschaft und Klimaschutz (BMWK) 
More Funding Information
This research was funded by the German Federal Ministry for Economic Affairs and Climate Action under grant number ZF4736301LP9.
Publication version
publishedVersion
Lizenz
https://creativecommons.org/licenses/by-nc-nd/4.0/
Thumbnail Image
Name

1-s2.0-S2212827122003997-main.pdf

Size

720.7 KB

Format

Adobe PDF