English: This image shows the effect of applying eight different filters on a test image by multiplying the spectra obtained with discrete cosine transform (DCT). Common effect such as low-pass and high-pass can be seen.
I created the image completely myself, using my own photo and Aileron font (CC0, by dot colon) and I release it under CC0. I used GIMP and GNU Octave. The Octave command used to filter the image is (consider CC0 as well):
pkg load signal; for i = 1:24, img = uint8(idct2(dct2(double(rgb2gray(imread("~/Documents/fft dct examples/test.png")))) .* rgb2gray(double(imread(strcat("~/Documents/fft dct examples/filter",num2str(i),".png")))))); imwrite(img,strcat("~/Documents/fft dct examples/result",num2str(i),".png")); end
The spectrum image is displayed directly without any smart mapping with:
imshow(uint8(dct2(rgb2gray(imread("~/Documents/fft dct examples/test.png")))))
Therefore the values are cropped and the spectrum image loses some information about the original image. It is there only to give an idea about what the spectrum looks like.