My grandma likes to sew. Years ago, she found a company who converted photographs into stitching plans. However, then that company disappeared; so I made her a little programme in Visual Basic (yes, really) that did the same. It wasn't fully automatic, though: It just "palettised" an image into a CSV file, which I then processed manually. That is:
- Convert the numeric palette into Wingdings of increasing density.
- Take counts of each colour and match them up to the palette. (This gives an estimate of how much wool is required.)
- Format the CSV nicely in Excel.
- Put everything together and print on several pieces of paper.
Note that a significant amount of preprocessing was also done to the image to put it into a "useful" state. Ordinarily, it's downsampled to eight shades of grey (as it's hard to find wool in other colours with a good range of graduations), but this has to be done sensitively: You need to strike a balance, in the dithering, between complexity (i.e., awkwardness to sew) and image quality. Also, it's worth pointing out that the resolution of such embroideries is pretty low; like 100×125 at about 12dpi low.
Anyway, did I mention how long ago this was? Well, it turns out that said VB programme doesn't work on modern platforms. Moreover, I have less time to do all the necessary pre- and postprocessing. However, I still want to make these things for my grandma: It gives her something to do and the output actually looks pretty cool.
So I'm remaking it, but in the hope of having it work (almost) fully automatically! I had planned on using Node.js, but I'm getting pretty sick of all that asynchronous nonsense and I've been teaching myself Haskell, which is a joy to an ex-mathematician. A Haskell implementation it is!