I'm pricing an FR5 faceplate inspection setup and our reject images include a bright line that slides sideways when the presentation changes. Everybody calls it a scratch. Nobody has the physical inspection result. Can we sort that cheaply before the next quote adds another light?
Have the acceptance owner examine the identified physical samples first; otherwise the lighting trial is being judged against labels that nobody can explain to the eventual operators.
Retain both labels in the history, but ask for one current judgement under defined viewing conditions before using that piece as a definite reject reference.
Did the same area stay in view when you changed presentation? A pair of images with identifiable surface features would help show what moved relative to the part.
That comparison can inform the imaging work, provided the sample identity and physical judgement remain attached to it rather than inferred from the bright line alone.
Same area, yes. The line crosses a small tooling feature between images. Quality is looking at the actual part with agreed viewing conditions. I haven't called it good, just stopped calling the photo proof of a scratch
We had a reference folder where uncertain parts quietly became rejects after a handover. Giving uncertain samples their own status helped. Include who resolves them, or they can sit there until somebody gets impatient and relabels the lot.
Quality owns the suspect list now. They've supplied a small set of agreed good and bad parts for the lighting comparison. This faceplate stays suspect, so no victory lap for the lamp yet
Use those agreed samples when documenting the operator inspection view too, so the eventual handover explains the acceptance standard rather than merely where to look for a bright pixel.