It's not all noise. There's a heavy Moire pattern due the fact that you rasterized something already rasterized. One gets the same problem if he scans a printed photo. Digital signal prosessing mathematician would say that you have got a visible difference of the sampling frequencies both in the vertical direction and in the horizontal direction.
You should note that the pattern does not exist everywhere. That's caused by the lens distortion. There's a scaling differences between the areas of your image.
A precaution for future photos:
Take several photos at the same distance but having a little different zoom settings. Use a small aperture, for example 1:11 to keep the barrel distortion low. Also have a good tripod to be able to keep the ISO as low as possible for low real noise and no camera shaking blur.
If you want to fix your already taken photos, you can use some good noise eliminating program. The following example is the output from Neat Image (the 5 years old free version). The noise sample was taken from the dense pattern in the middle of the photo.

Unfortunately all faint buttont texts (Save, help, Exit) have nearly vanished. The remedy was to copy the cleaned image over the original as a new layer. Then the nearly vanished parts were erased from the cleaned version.
The cleaned layer got finally the level adjustment which was suggested in earlier answers. Now it works because the unwanted moire pattern remnants really are lighter than the wanted texts and shapes. See a screenshot of that final step.

I don't believe you can do much better by thislike simple methods except by redrawing.
Lighten only
mode. The GMIC plugin also has a "Median" filter that could work wonders. – xenoid Apr 27 '17 at 16:19