A fixed camera, small presentation changes, very different streaks (planning camera inspection of reflective parts)

RosaAbbott0031 · 27 Aug 2026, 14:27 UTC

Reply to discussion
RO
RosaAbbott0031
Our planned planning camera inspection of reflective parts in a precision metal-parts inspection area has Fairino FR5 presenting a polished metal cover to a fixed camera. The example images contain bright streaks that shift with small presentation differences, yet those streaks are being labelled as scratches. Before commissioning, I'd like to establish reliable references and image conditions for threshold tuning.

19 replies

JA
JaneBarnes0597
Replying to RosaAbbott0031

@RosaAbbott0031 Check whether the suspected scratches were confirmed on physical samples under the agreed appearance procedure. Link each image to its sample identity and inspection result first.

14 points
RO
RosaAbbott0031
Replying to JaneBarnes0597

@JaneBarnes0597 Some labels came from the pictures alone. We still have the samples, but those labels aren't tied to physical inspection results. That's irritatingly circular.

11 points
JA
JaneBarnes0597
Replying to RosaAbbott0031

Keep the image-only labels uncertain pending physical inspection. A controlled comparison of lighting and presentation can then investigate whether the moving streaks indicate reflection effects.

17 points
JO
JonasAli0198
Replying to JaneBarnes0597

A scratch's visibility can change with presentation too. Don't let 'moving streak' quietly become the rule for relabelling something acceptable.

-8 points
JA
JaneBarnes0597
Replying to JonasAli0198

@JonasAli0198 Correct. I meant a reason to investigate the imaging, not a new defect label. Physical acceptance stays with the agreed inspection method.

22 points
LI
LiamCarter1012
Replying to JaneBarnes0597

@JaneBarnes0597 Should uncertain images be treated as acceptable, or kept somewhere separate while their status is established?

15 points
JA
JaneBarnes0597
Replying to LiamCarter1012

Retain their sample identities and undecided status. They shouldn't become confirmed acceptable or defective examples while the physical review remains unresolved.

25 points
RO
RosaAbbott0031
Replying to JaneBarnes0597

@JaneBarnes0597 Our two comparison folders changed exposure and presentation together. I can't tell which change made the images look better. That comparison needs a caveat too.

18 points
CH
ChenChan1051
Replying to RosaAbbott0031

I had a separate inspection setup where the 'better lighting' folder used easier samples too. Its name implied a lighting conclusion that the comparison didn't support.

16 points
RO
RosaAbbott0031
Replying to ChenChan1051

@ChenChan1051 What did you keep beside each image? I can manage sample ID and settings, but I don't want a form nobody fills in.

9 points
CH
ChenChan1051
Replying to RosaAbbott0031

For my separate setup, sample identity, exposure, lighting arrangement and presentation condition were the useful essentials. They let us understand comparisons without relying on recollection.

13 points
JA
JaneBarnes0597
Replying to ChenChan1051

Those records can support the vision specialist's comparison of image conditions. Holding the representative sample set constant will make an acquisition change easier to assess.

6 points
LI
LiamCarter1012
Replying to JaneBarnes0597

Would a learned model just cope with the lighting differences?

14 points
JA
JaneBarnes0597
Replying to LiamCarter1012

Any learned model still depends on reliable labels and representative evaluation. Changing the method doesn't resolve unconfirmed physical defects or an uncontrolled comparison of image conditions.

23 points
JO
JonasAli0198
Replying to JaneBarnes0597

Keep physical sample identity in mind when forming evaluation sets. Different images of a sample used for tuning don't provide an independent sample-level evaluation.

14 points
RO
RosaAbbott0031
Replying to JonasAli0198

@JonasAli0198 I'll keep all images of one sample together when we split tuning and evaluation sets. Our file names alone wouldn't have caught that overlap.

8 points
JA
JaneBarnes0597
Replying to RosaAbbott0031

@RosaAbbott0031 Use the held-out samples to report defect misses and false rejections independently. A combined agreement figure can conceal the type of error that matters to the process.

24 points
RO
RosaAbbott0031
Replying to JaneBarnes0597

@JaneBarnes0597 The reference problem is clearer, but I still can't judge inspection performance from our example images. The physical labels and image comparisons need more work.

13 points
JA
JaneBarnes0597
Replying to RosaAbbott0031

The remaining physical-label and comparison work explains the open judgement. Clearer uncertainty doesn't yet establish inspection performance.

14 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.