Once again, "Lighting Theory".

The purpose of imaging inspections is almost always aimed at replacing visual inspections.
Visual inspection is most often performed under general room light or bright fluorescent light.
However, the illuminators used for image inspection are often LEDs.
It is said as if it is common sense to use LEDs to increase the contrast of defects for imaging.
This is the pitfall.
It only increases the contrast of "some defects," not "all defects.
The result of "devising" such as single wavelength or directional imaging is an image that is far removed from what we are accustomed to seeing visually, and there is no correlation with visual inspection.
It is fundamentally dark. The contrast does not increase because it is dark.
The reason why the image appears to glow sharply is because the directivity of the camera and the directivity of the illumination happen to coincide. If the camera is tilted a little, the image becomes dark instantly. In the case of visual inspection, it is easy to find the object because it "glows sharply for a moment" by moving it with the hand. This is not because the "human eye" is great, but only because it is achieved in combination with the "hand.
Ultimately, lighting is,
1) An illumination system similar to general room light
2) Coaxial falling slopes with weak directivity
is almost always sufficient.
If either of these is not good enough, it means that stable images cannot be taken, and it is not suitable for image inspection.
To be honest, I have not yet found an illuminator that is "the one".
However, I have been getting ideas for making one, and I am pondering whether I should make it myself or outsource its manufacture to a manufacturer.