Our example images keep turning shiny patches into defects - inspecting reflective surfaces for scratches

LinAllen0270 · 20 Jul 2026, 21:05 UTC

Reply to discussion
LI
LinAllen0270
Our planned inspecting reflective surfaces for scratches in a small finished-parts production cell has Fairino FR10 presenting a brushed stainless trim plate 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.

15 replies

AL
AlexChan1081
Replying to LinAllen0270

Were those scratches confirmed on the physical samples using your agreed appearance method? Start by linking each image to a sample and its inspection result.

16 points
LI
LinAllen0270
Replying to AlexChan1081

I've found that some labels were assigned from the images themselves. The samples are still available, but those labels have no corresponding physical inspection results, making the reference circular.

2 points
AL
AlexChan1081
Replying to LinAllen0270

Mark those labels uncertain and have the samples inspected properly. Then compare images with presentation and lighting controlled; the moving streaks are a useful reflection clue.

24 points
OM
OmarBell0665
Replying to AlexChan1081

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

11 points
AL
AlexChan1081
Replying to OmarBell0665

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

23 points
LE
LeahAli0234
Replying to AlexChan1081

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

19 points
AL
AlexChan1081
Replying to LeahAli0234

@LeahAli0234 Keep them explicitly undecided, with the sample link. Don't use them as confirmed examples of either class until the physical review settles them.

9 points
LI
LinAllen0270
Replying to AlexChan1081

I've checked the image folders and found that exposure and presentation both changed between them. Their improved appearance can't be attributed to either change individually.

2 points
JA
JackBell0689
Replying to LinAllen0270

@LinAllen0270 On my inspection setup, a folder called 'better lighting' also contained easier samples. The folder name was doing a lot of unearned work.

14 points
LI
LinAllen0270
Replying to JackBell0689

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

1 points
JA
JackBell0689
Replying to LinAllen0270

On my setup we kept sample ID, exposure, lighting arrangement and presentation condition. Enough to explain a comparison without reconstructing the whole session from memory.

5 points
AL
AlexChan1081
Replying to JackBell0689

Use those details to ask your vision specialist for controlled comparisons on representative surfaces. Keep the sample set fixed when assessing an acquisition change.

22 points
LE
LeahAli0234
Replying to AlexChan1081

Would a learned model just cope with the lighting differences?

12 points
AL
AlexChan1081
Replying to LeahAli0234

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.

13 points
OM
OmarBell0665
Replying to AlexChan1081

@AlexChan1081 And don't tune and test on different pictures of the same physical sample. That isn't the independence the result would suggest.

18 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.