We're planning inspecting reflective surfaces for scratches, with Universal Robots UR5e presenting a polished metal cover to a fixed camera in a precision metal-parts inspection area. Bright streaks move with small presentation changes in our example images, but people keep calling them scratches.
We haven't commissioned the cell. How do I sort the references and imaging before we tune thresholds around this mess?
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.
@BrunoChen1148 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.
@BethBrooks0846 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.
I've settled the review method: physical reference checks plus controlled image comparisons. That answers where to start; whether our inspection concept is suitable still depends on the resulting evaluation.