Moiré pattern - avoiding or removing

[JuergenB.]

Dear Community, my question has its origin in machine vision, but it should fit here. I use a 1 chip camera to make photos of flat objects with fine structures. These photos have to be maximally sharp. Sometimes there are structures so small, that the Nyquist–Shannon sampling theorem is violated and I get strong Moiré patterns. In the worst case I get a rainbow in an area, which is actually just a tight black and white grid. Of course the effect would go away if I bring the lens a bit out of focus, so that is filters the high frequencies out, but that is not an option for me because I need the maximally sharp edges. Do you know another method to avoid this effect? Regards JuergenB.

 

[JuergenB.]

I switched to a trilinear camera (its all about linescanning in my case but the effect is the same as with area scan). With the correct line frequency it is as good as a 3-chip-camera. Just tested it and the problem is solved.