node-sanjuuni

Converts images and videos into a format suitable for ComputerCraft, based on sanjuuni.

Usage

Import this library using require:

const sanjuuni = require("sanjuuni");

You may call initOpenCL before doing any conversion to use GPU acceleration via OpenCL if available. Note that OpenCL is not currently built by default, but power users may recompile the module to link in OpenCL.

Image conversion generally has four (or five including source decoding) steps:

1. Load the raw pixel data into an RGBImage instance.
2. Generate an optimized palette using a reducePalette function.
3. Convert the image to an indexed image using thresholdImage or a ditherImage function.
4. Generate the final output using one of the make* functions.

Alternatively, using CIELab color space for more optimal color accuracy:

1. Load the raw pixel data into an RGBImage instance.
2. Convert the image into Lab space with makeLabImage.
3. Generate an optimized palette using a reducePalette function.
4. Convert the image to an indexed image using thresholdImage or a ditherImage function.
5. Convert the palette back into RGB space with convertLabPalette.
6. Generate the final output using one of the make* functions.

For example, to generate a Lua script using Lab color, k-means quantization and Floyd-Steinberg dithering, operating on a BGRA buffer of pixels:

const img = sanjuuni.makeRGBImage(pixels, width, height, 'bgra');
const lab = sanjuuni.makeLabImage(img);
const labPalette = sanjuuni.reducePalette_kMeans(lab);
const idxImg = sanjuuni.ditherImage_floydSteinberg(lab, labPalette);
const palette = sanjuuni.convertLabPalette(labPalette);
const luaFile = sanjuuni.makeLuaFile(idxImg, palette);

Note that this module does not have any built-in image decoding capabilities; use other modules to decode files if necessary.

See the TypeScript typing file index.d.ts for complete documentation on the available functions.

License

node-sanjuuni is licensed under the GPLv2 license (or later at your choice).