Here are some simple animations to make it easier to understand the connection between repeating patterns in an image and the resulting fourier transform:
Blur (high frequencies are further away from the center, as you can see when we remove them by blurring the image):
Rotate image (as you can see, the angle of image features is the same in the frequency domain, yet we always have a horizontal/vertical lines through each "star"):
Rotate pattern (only the pattern is moving; as you can see the pattern is clearly defined by the bright stars while the rest is static):
Scale image (this effectively increase/decreases the frequencies so the "stars" move in/out):
Scale pattern (same, but only pattern - so the static areas are thus the base image):
Translate pattern sideways (notice that the frequency and amplitude remains ~fixed, and it's just the phase information on the right that is changing):