Cannot tune the FR5 trim-plate check from unconfirmed bright streaks

GabrielCarter0975 · 28 Jun 2026, 16:10 UTC

Reply to discussion
GA
GabrielCarter0975
Our proposed camera station presents brushed stainless trim plates with an FR5. Small presentation changes move the bright streaks, but the image folder calls them scratches. I need a defensible physical reference and controlled comparison before these captions become our threshold targets.

11 replies

LE
LeoChen1145
Replying to GabrielCarter0975

Keep the plates identified and have quality decide the actual marks under an agreed appearance method, then compare the camera views with those decisions; moving light alone can't tell you which surface is acceptable

21 points
GA
GabrielCarter0975
Replying to LeoChen1145

The plates are retained and individually numbered. Quality has not approved the captions; they came from our image discussion. I will keep them as disputed observations, not reference decisions.

19 points
GA
GabrielCarter0975
Replying to GabrielCarter0975

Quality identified one shallow scratch and accepted the other two streaked regions as finish variation. The physical locations are now marked on reference photographs. Our next comparison needs to show whether the actual scratch remains detectable when presentation varies.

16 points
AN
AnilBaker0462
Replying to GabrielCarter0975

When the same plate is removed and replaced, does the observed seating change, or are you deliberately changing its angle? Those require different comparisons.

1 points
LU
LucaBennett0700
Replying to GabrielCarter0975

And keep the accepted examples in that comparison. Otherwise the cheapest way to find the scratch could be to reject every brushed plate, which is not much of an inspection system.

8 points
LE
LeoChen1145
Replying to AnilBaker0462

Anil's question comes first for me. A stable seat gives you a useful baseline; then the permitted presentation range can be tested deliberately, instead of mixing loader variation and lighting changes in one folder

14 points
GA
GabrielCarter0975
Replying to AnilBaker0462

Both were mixed in the original images, Anil. We repeated placement without changing lighting and found the holder allowed one edge to lift. We have not yet tested deliberate angle changes from a repeatable seat.

22 points
AN
AnilBaker0462
Replying to GabrielCarter0975

Then record which edge lifted and the intended contact arrangement for the fixture designer. A brighter lamp will not explain why the plate is sitting differently each time.

9 points
IM
ImranAdams0131
Replying to GabrielCarter0975

In our image review, crops stayed linked to their originals. Keep those links here too; a close-up of the scratch can hide the lifted edge explaining the presentation.

4 points
GA
GabrielCarter0975
Replying to ImranAdams0131

Agreed, Imran. Originals and settings are retained, with crops only as linked views. The designer has the observed lift and intended contacts; holder correction and the controlled detection comparison are both still outstanding.

5 points
LE
LeoChen1145
Replying to GabrielCarter0975

Once the seat is repeatable, include the accepted finish regions as well as that scratch at the intended presentation limits; the physical reference decisions are useful progress, not yet evidence that this camera arrangement separates them

15 points

Add to the discussion

Welcome to Application Robot

Everyone can read the forum. Sign in or create an account to start a discussion, reply, or upload photos.

Forgot your password?

By creating an account, you agree to our Terms and Conditions and community guidelines. Read our Privacy Policy for how your information is handled.