I'd been looking for a way to convert images to tables and it seems this is doable using the "irfanview" program by saving an image as a ppm in binairy data 
So right now I've got an ifft-oscillator running, that gets it's fft values by scanning through RGB-values of the image
So, to do this:
1. get an interesting image somewhere
2. use irfanview to "select a part and cut" or scale it to a nice size (say 512x128 pixels)
3. use color corrections and effects to adjust image. Turn up the brightness at least a bit to ensure black ain't really black, as this will leave a blank space instead of a value of zero, it seems, making the array shorter then it supposed to be.
4. save it as a .ppm file and set it to binairy in the popup file-settings window. After saving, open the file with notepad and remove the header (irfanview blabla, plus the extension of the size (this data ends with the width and height of the image, just remove this to only keep the RGB data. It won't be readable anymore by any program on your computer, but the axoloti still can)
5. put it on your SD-card and just load it with the factory table load module with extension .ppm (where you'ld normally put something like .tab).
6. depending on whether you save as .ppm, .pgm, .pbm:
.ppm-> RGB pixels: table will give you 3 values in a row for each pixel, which are RED GREEN BLUE values from 0 to 255. then repeats for next pixels like: 46 63 83 66 74 46 34 10 244 etc), so an image of 64x64 will have 64x64x3 values in its table.
.pgm grayscale image version, haven't tried it yet, but guess it's just the greyscale values one by one for each pixel.
.pbm bitwise image, just zero's and ones.
also, another program seems to be very nice to convert table values between hex, string and binary:
https://codebeautify.org/string-binary-converter