Paper 2022/924
FICS PCB X-ray: A dataset for automated printed circuit board inter-layers inspection
Abstract
Advancements in computer vision and machine learning breakthroughs over the years have paved the way for automated X-ray inspection (AXI) of printed circuit boards (PCBs). However, there is no standard dataset to verify the capabilities and limitations of such advancements in practice due to the lack of publicly available datasets for PCB X-ray inspection. Furthermore, there is a lack of diverse PCB X-ray datasets that encompass images from X-ray Computed Tomography (CT). To address the lack of data, we developed the first comprehensive publicly available dataset, "FICS PCB X-ray," to aid in the development of robust PCB-AXI methodologies. The dataset consists of diverse images from the tomographic image domain, along with challenging cases of unaligned, raw X-ray data of PCBs. Further, the dataset contains projection data and the reconstructed volume which is converted into a Tiff stack. Annotated X-ray layer images are also available for image processing and machine learning tasks. This paper summarizes the existing databases and their limitations, and proposes a new dataset, "FICS PCB X-ray''.
Note: Dataset Release of "FICS PCB X-ray", version 1.
Metadata
- Available format(s)
- Category
- Applications
- Publication info
- Preprint.
- Keywords
- AXI dataset PCB dataset image processing computer vision machine learning automated X-ray inspection.
- Contact author(s)
-
dhwanimehta @ ufl edu
jtrue15 @ ufl edu
paradiso @ ufl edu
njessurun @ ufl edu
dwoodard @ ece ufl edu
nasadi @ ece ufl edu
tehranipoor @ ece ufl edu - History
- 2022-07-15: revised
- 2022-07-15: received
- See all versions
- Short URL
- https://ia.cr/2022/924
- License
-
CC BY
BibTeX
@misc{cryptoeprint:2022/924, author = {Dhwani Mehta and John True and Olivia P. Dizon-Paradis and Nathan Jessurun and Damon L. Woodard and Navid Asadizanjani and Mark Tehranipoor}, title = {{FICS} {PCB} X-ray: A dataset for automated printed circuit board inter-layers inspection}, howpublished = {Cryptology {ePrint} Archive, Paper 2022/924}, year = {2022}, url = {https://eprint.iacr.org/2022/924} }