1. Re-sizing subsamples the image so a transition which covers a few pixels will still have a pixel or two after re-sizing where as a single pixel transition will disappear all-together. So if you have to sharpen an image, do it after you resize it. If you have a razor sharp image you want to resize, bleary it up slightly first. Resize it, then resharpen it.
2. The less area a transition occupies in an image, the sharper we think it looks. Thus, the sharpening algorithm looks for a transition and then increases its amplitude while reducing its area. So the more transition you've got to start with the better will be the result. If an image is flat with almost no transitions, increase the difference between its light and dark sections (either by increasing the contrast or by gamma correcting) before sharpening it.
3. A transition covering a large area (number of pixels) will take a large number of iterations of the sharpening algorithm to reduce the area of the transition where as one covering a small area will get sharper faster. So resize an image before you sharpen it.
One of the main problems I encountered when observing and taking pictures of the Moon, is its brightness. As it has no atmosphere, it is illuminated by the full power of the Sun. It has a reasonably high reflectance factor and we usually observe it on a clear night when it is high in the sky, thus reducing the loss of brightness, due to our atmosphere, to a minimum. This means it looks about as bright as a sunny summer day at the beach.
Unfortunately, to our eyes, it appears very small and the auto exposure system of our eyes/brain does not correctly allow for this high brightness. We see the moon in an over-exposed state. Similarly, cameras with auto exposure also see the moon in an over-exposed state and usually it turns out as a white blob.
In order to see the moon more clearly we usually view it through a telescope. The smallest one, that is recommended for serious viewing, has a front element diameter of 3 inches (70 mm). If we take the fully open iris of an eye to be 7 mm, then the telescope gathers about 100 times the amount of light your eye gathers. Most telescopes use a magnifying eye-piece and to get the moon filling the eye-piece, you need about 50 times magnification. Thus every bit of the moon image presented to you in the eye-piece is twice as bright as what you can see with your naked eye. Of course, peering through the eye-piece you'd be lucky to fill 20% of your field of view so the 'scope is effectively outputting about 10 times too much light. Your auto exposure cannot handle this due to the small area the moon occupies on your retina so the image over-exposes and becomes a very bright flat image. Eventually your eye will water with the distress. The high brightness also gives loss of contrast by causing flare in the atmosphere and lenses, and internal reflections in the telescope tube.
These problems are partially solved by a Moon-filter. This is a neutral density filter, placed behind the eye-piece, which passes about one sixth to one twelveth of the light. This does not fix the flare, though.
Camera images suffer a similar fate, especially cheap ones.
Another problem with holding a camera up to the eye-piece, is focus. Normally you look through the eye-piece and focus the telescope as sharp as possible for your eye. You then put the camera to the eye-piece and hope for the best. You dont get the best because the camera is focussed at infinity but your eye is usually not, hence the 'scope is not focussed for the camera and your shots come out slightly bleary (soft, out of focus). One way to lessen this problem is to train your eye to relax when focussing the telescope. This way the 'scope will be focussed closer to infinity which will help the camera focus.
The pictures below were taken with a cheapo digital camera which had auto exposure and a tiny fixed focus lens set back in the body. It was placed up close to, but not touching, the 20 mm eye-piece of a friend's 3 inch x 3 ft refracting telescope which unfortuneatly had no Moon-filter. The eye-piece exit pupil was about 1.6 mm dia. and the camera lens was about 3 mm dia. It was near-on impossible to keep the lenses lined up, as well as tracking the 'scope by hand, and also allow for the fact that the camera took the picture about a third of a second after you pressed the button. So in the end we had to take about 100 pix to get around 5 usable ones. The following composite contains a few samples of those typically bad results.