I want the labels checked before the proposed FR5 camera trial goes further. The broad stripe marked scratch belongs to a faceplate quality has accepted. It changes position when the plate is tilted.
We have the retained parts and original images. One rejected faceplate is in the box too, but its number doesn't appear anywhere in the image filenames. The folder looks ready for training until you try to match it to the metal.
There is a sheet, but the rejected plate was entered under a temporary sample number. I've found that number in the originals. Quality is checking the feature against the actual plate.
And the accepted one? Don't quietly lose it because its reflection makes the folder inconvenient. That is exactly the surface the proposed camera has to tolerate.
I would put the temporary and permanent numbers together while you rebuild the set. We once relabelled samples beautifully and left the original image sheet using the old names. Next person assumed half the evidence had disappeared. A tiny table would have spared a very boring afternoon.
Quality matched the reject and confirmed the small mark near its screw hole. The accepted plate's stripe has no corresponding rejected feature. I've linked both sample numbers and removed the guessed scratch label, not the accepted image.
Sara, the holder leaves it visible, but the bright band crosses it in the current view. Hassan, the tutor has the corrected set with the number cross-reference. Imaging still needs work.
Yes. Vision is comparing them under the same located presentation, then an alternative lighting arrangement. We are not changing the labels to make either arrangement look better. Thanks, Jack, the tutor immediately asked about the old number.