The idea I present here was partly inspired by David Byrne's artwork from the cover of an early Talking Heads album -- he took a lot of close-up polaroids of the band members, and assembled those photos into much larger images of the band.
I suffer from OCD -- I hate to throw away anything that might have some sort of "re-usability" (in olden times I would've been called a "pack rat" but nowadays it's called OCD ;-) ). I used to live where all cardboard was recycled; then I moved where they don't recycle the "coated" stuff (cereal boxes, etc.); all the cardboard packaging that has that smooth coating on it, I'm supposed to throw in the trash. I hate to do that -- seems so wasteful. So, all these cardboard boxes, I flatten 'em & stuff 'em in the basement. What to do with 'em all?
Well, how about this: cut them into uniform squares; scan them into the computer, carefully numbering, organizing, and storing them (explication forthcoming)... use the resulting collection of scanned images as input to a photomosaic program. Take an open source photomosaic program, and modify it so its output lets me identify each piece of cardboard (that's why I need to organize 'em, so I can easily retrieve any given square)... then I can recreate the output photomosaic in "meatspace."
Step 1: Materials List
1. scanner (digital camera can work in a pinch)
2. colorful cardboard (save the boxes for stuff you're already buying anyway; you'll have more than you'd expect in no time)
3. laser cutter (or paper cutter, or scissors, or x-acto knife, or...)
4. An open-source photomosaic program (I like Gijs Molenaar's mozaic.py)
5. coding skills to modify the program (that's why it needs to be open-source)
Step 2: Make the Cardboard Pieces Uniform in Size
A laser cutter would be perfect for this job (cutting the cardboard into squares), but a paper cutter'll do in a pinch (or even scissors, a boxcutter, x-acto, etc.).
If you want to use smaller "tiles," I highly recommend scanning them before cutting them up (see the image (sorry it's sideways)). Then, use a program (suggestion: Python makes it pretty easy to write -- yay, free imaging libraries!) to subdivide the scanned image, and use the (laser or paper) cutter to subdivide the cardboard. In the photo, I've marked each tile with two numbers: one to indicate which scan it's part of, the other to identify each individual 1" square. This is necessary to be able to assemble the mosaic. You also want to store the tiles, after they're scanned, using some sort of scheme to allow easy retrieval by tile ID (in this example, that'd be scan#+tile#). My suggestion: stack 'em ten-tiles-high, in rows (of ten, perhaps) & columns... somewhere your kitten/puppy/toddler won't mess with 'em.
For example, you could write your Python program to take parameters for the number of horizontal & vertical subdivisions, & the "scan #":
python subdivide.py input.jpg 8 8 2
...could mean split the image "input.jpg" into 64 squares, and create output files named like:
...where the "2" in "scan2" just means: this is the second square of cardboard I scanned (and the "scan" part isn't really needed, actually).
Step 3: That's Pretty Much It for Now
Consider: how many 8" square cardboard tiles? Suppose you can find an area about 8' by 8' (a little bigger) in which to stack your tiles. At 10 tiles per pile, that'd be 144 piles, or 1440 tiles; if composed into a square mosaic, it'd measure 80 feet on a side. If you cut & scanned, say, 10 a day, you could create that many tiles in less than half a year. If 4 people did as much, the mosaic could be 160 feet on a side; 16 people, 320 feet, and now we're getting to the point where the image might be recognizable in a satellite photo... I think (I'm not quite sure, actually). Btw, if anyone spots errors in my math, please leave a comment.
I hope this might be enough for someone to get inspired by... if there's anyone else as eccentric as I am out there... I have LOTS of colorful cardboard, and would be happy to deliver it in the northeast US if anyone wants it... :-)