A systematic literature review of the application of artificial image data for visual defect detection
Publication date
2025-09-29
Document type
Übersichtsartikel, Überblicksdarstellung
Author
Organisational unit
Publisher
IEEE
Series or journal
IEEE access
ISSN
Periodical volume
13
First page
172674
Last page
172691
Peer-reviewed
✅
Part of the university bibliography
✅
Funding(s)
Language
English
Keyword
Defect detection
Deep learning
Artificial image
Synthetic image
Data augmentation
Generative adversarial network
Defect recognition
Systematic literature review
Abstract
Artificial images for defect detection have gained growing importance in the recently developed defect detection architectures. This systematic literature review examines the use of artificial image data for visual defect detection. It also provides an overview of the methods used for this purpose. Following PRISMA guidelines, the review analyses existing literature to identify key areas where artificial image generation is being applied or could be implemented to enhance defect detection capabilities. It was outlined that data augmentation is currently the most commonly used image generation method to improve individual datasets for defect detection tasks. The results provide insights into the current state of the art and potential future directions for addressing data scarcity in visual defect detection. The knowledge gained in this study can be used by researchers and users to find suitable methods for the desired field of application in order to overcome the problem of data scarcity, and the findings also show where there is still a need for research.
Description
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
Version
Published version
Access right on openHSU
Metadata only access