DOCSEARCHbyAlgolia
Submit Docs

quantize

repository·master·Indexed 41 days ago

https://github.com/olivierlesnicki/quantize

A Node.js module for color quantization based on Leptonica. It implements the Modified Median Cut Quantization (MMCQ) algorithm to reduce a large set of RGB pixels into a smaller, optimized color palette of 2 to 256 colors.

Tokens
1K
Snippets
5
Records
6
Agent score
50%

What's inside quantize

  2. Perform color quantization with quantize()

    master

    To quantize a set of colors, require the module and call the quantize function with two arguments:

    1. arrayOfPixels: An array of pixels, where each pixel is represented as an [R, G, B] array.
    2. maximumColorCount: The maximum number of colors allowed in the resulting reduced palette.

    The function returns a colorMap object used to access the palette and map individual pixels.

    var quantize = require('quantize');

var arrayOfPixels = [[190,197,190], [202,204,200], [207,214,210], [211,214,211], [205,207,207]];
var maximumColorCount = 4;

var colorMap = quantize(arrayOfPixels, maximumColorCount);

  3. Retrieve the reduced color palette using .palette()

    master

    After creating a colorMap via quantize(), call the .palette() method to get the array of reduced colors.

    // Returns the reduced palette as an array of [R, G, B] arrays
colorMap.palette(); 
// Example output: [[204, 204, 204], [208,212,212], [188,196,188], [212,204,196]]

  4. Map a pixel to the reduced palette using .map()

    master

    To find the closest color in the reduced palette for a specific pixel, use the .map(pixel) method on the colorMap object. Pass an [R, G, B] array as the argument.

    // Returns the reduced pixel color
colorMap.map(arrayOfPixels[0]);
// Example output: [188,196,188]

  5. Quantize colors using MMCQ.quantize()

    master

    The quantize function implements the Modified Median Cut Quantization (MMCQ) algorithm. It takes an array of pixels and reduces them to a specified number of colors, returning a color map object used to retrieve the new palette and map original pixels to the reduced set.

    Parameters:

    • pixels: An array of pixels, where each pixel is represented as an [R, G, B] array (e.g., [[190, 197, 190], [202, 204, 200]]).
    • maxcolors: The target number of colors for the palette. Must be between 2 and 256 inclusive.

    Returns:

    • A CMap object (color map) if successful.
    • false if the input is invalid (e.g., pixels is empty, or maxcolors is out of range).

    Usage Example:

    var myPixels = [[190,197,190], [202,204,200], [207,214,210], [211,214,211], [205,207,207]];
var maxColors = 4;

var cmap = MMCQ.quantize(myPixels, maxColors);

if (cmap) {
    var newPalette = cmap.palette(); // Returns the reduced palette
    var newPixels = myPixels.map(function(p) { 
        return cmap.map(p); // Maps original pixel to the nearest palette color
    });
}
    var myPixels = [[190,197,190], [202,204,200], [207,214,210], [211,214,211], [205,207,207]];
var maxColors = 4;

var cmap = MMCQ.quantize(myPixels, maxColors);
var newPalette = cmap.palette();
var newPixels = myPixels.map(function(p) { 
    return cmap.map(p); 
});

  6. Use the CMap object to manage color palettes

    master

    The CMap object returned by quantize() provides methods to interact with the generated color palette and perform color mapping.

    Methods:

    • palette(): Returns an array of the colors in the reduced palette. Each color is an [R, G, B] array.
    • map(color): Takes an [R, G, B] color and returns the closest color from the generated palette. It first attempts to find a match within the color space boxes; if no exact box contains the color, it falls back to the nearest() color.
    • nearest(color): Calculates the Euclidean distance between the provided [R, G, B] color and all colors in the palette, returning the one with the smallest distance.
    • size(): Returns the number of colors in the palette.