I'm sorting faceplate images for a proposed FR10 inspection cell, and the brightest 'scratch' changes position between views of the same part. Nobody kept the physical judgement with the image labels. I'd like to salvage useful references before threshold tuning turns this argument into software.
Can you still match those views to the actual faceplate, and have quality judge its condition using the agreed product criteria before any image gets a defect label?
For that pair, yes. There's a sample number in the uncropped originals. Some older pictures have no identity anywhere, so I'm putting those aside for now rather than inventing one.
Keep the unidentified pictures as examples of the imaging problem if useful, but they cannot support a traceable acceptance reference in their present form.
Our inspection team had a similar disagreement over a polished ring. The surface was real; the apparent line was mostly lighting. We needed the physical inspection and the imaging experiment recorded separately, or each group kept thinking the other had settled the question.
That distinction matters here too; a changing highlight demonstrates presentation sensitivity, but does not prove the faceplate has no genuine defect elsewhere.
A controlled small comparison would help define what the larger collection must record, including illumination, camera settings and sample presentation.
We've reproduced it at the bench with the same faceplate and a small orientation change. Quality also found a real fine scratch that is visible in only some views. Both effects need handling.
It's identified and retained now. I'm adding the physical inspection result beside the paired images, with the setup details. The uncertain labels are coming out of the training folder.
Coming out of training, yes, but keep a review set too. Sorry, I mean retain the uncertainty somewhere it can be revisited rather than deleting the troublesome pictures.
Use an explicit unresolved judgement for those examples; forcing them into the accepted or rejected class would make the dataset appear more certain than the inspection team is.
Assign that responsibility alongside the imaging setup ownership, so a lighting adjustment does not leave old reference images silently representing a different inspection condition.
Quality will own judgements; the vision lead will own acquisition settings. We have a small traceable review set now, not an approved production classifier. The old screenshots are retained outside training with their limitations.
That gives you a defensible starting point; the next question is whether the chosen imaging conditions reliably reveal the defects your product criteria require you to find.