フーリエ変換

本日フーリエ変換の使い道に関して読んでいます。
While I am particularly eager to see how this applies to financial modeling, I have opted to begin by seeing how the transform is applied to image processing. My motive for this choice is that image processing contains concrete input and output, so the applications should be plain to observe.
Interestingly, starting with an image of black and white stripes, the "frequency domain" output results in a solid black background marked with three white points in a horizontal line through the center. The center point is the DC frequency, F(0,0), which is apparently the "average brightness" of all points in the image. The other two points are the F(1/4, 0) and F(-1/4, 0) it seems, as the stripes are two pixels wide. I am eager to see how obtaining these frequency domain results allows for transformation back to the spatial domain with relevant alterations.