Suppose a photograph with text and numbers. I want to manage it in my editor with tools such as grep, standard text-processing things such as Vim's block-highlighting and also more advanced things such as Gimp's magic-wand-style pattern highlighting. How?
Analysis
This puzzle breaks down at least to parts
- OCR -- character recognization
- DSP -- proximity -algos, all puzzles not yet known
- Ascii-art -- creating fillers and "decorators" (not technical term)
For the sake of simplicity, suppose the writing is line-wise so you do not need to consider reprocessing multiple-lined documents. There is some working-prototype that does pretty good job in small scale with LaTex here, the multiple lined problem follows later.
but unfortunately, I left no comment, how I produced it. Since I know the foto, where it origins from, it could be gimp-plugin too. Or I reduced the colors drastically to 4, 8 or 16 and reduced the resolution to something screen-friendly like 80x40, and worked with a programming language along the pixels, writing different characters for different colors, which I then wrapped with ANSI codes to produce the colors, and replaced signs with words. Maybe partly by hand, if I remember correctly.