6 releases (breaking)

0.7.0 Jan 25, 2021
0.6.0 Apr 13, 2020
0.5.0 Apr 7, 2020
0.4.0 Mar 22, 2020
0.2.0 Oct 29, 2019

#328 in WebAssembly


Used in 2 crates

MIT license

23KB
443 lines

🎨 Pigmnts

Pigmnts is a library to create a color palette from an image built using Rust, compiled to WebAssembly. This allows for super-fast color palette extraction from an image on the web. It uses the K-means++ clustering algorithm to select the most commonly occuring colors from the image.

Examples with Web Assembly

As a JavaScript module

<script type="module">
  // Import the functions from CDN
  import init, { pigments } from 'https://unpkg.com/pigmnts/pigmnts.js';

  async function run() {
    /**
     * Load the wasm file.
     * Replace the URL with a different path if you want to use
     * self hosted wasm file 
     */
    await init('https://unpkg.com/pigmnts/pigmnts_bg.wasm');

    // Canvas element from which palette should be created
    const canvas = document.querySelector('canvas');

    // Call the pigments wasm function
    const palette = pigments(canvas, 5);
  }
  run();

</script>

In Node.js

  1. Start by installing the npm package
npm install pigmnts
  1. Configure your build process to copy the wasm file in the your build directory.

  2. Use it in code

import init, { pigments } from 'pigmnts';

async function run() {
  // Load the wasm file from path where wasm file was copied by the bundler
  await init('<path to the wasm file>');

  // Canvas element from which palette should be created
  const canvas = document.querySelector('canvas');

  // Call the pigments wasm function
  const palette = pigments(canvas, 5);
}
run();

Functions

Pigmnts exposes following function in WebAssembly

pigments(canvas: HtmlCanvasElement, k: number, mood: Mood|number, batch_size: number)

Arguments
  • canvas canvas element which has the image to be processed. Internally, the pixel data is taken from the canvas, and then clustered to create the color palette.
  • k defines the number of colors to be gathered from the image.
  • mood defines the weight function to use. Only 'dominant' mood is supported which has a value of 0
  • batch_size (optional) defines the number of pixels to randomly sample from the image. It should be greater than the total number of pixels in the image and the k. By default, all the pixels in the image are processed.
Return

Returns an Array of Objects where each Object is a color of the following format.

[
  {
    dominance: 0.565    // Dominance of color in image(From 0 to 1)
    hex: '#6DDAD0'      // 6-digit Hex color code
    rgb: {              // Equivalent RGB color
      r: 109,
      g: 218,
      b: 208
    },
    hsl: {             // Equivalent HSL color (Normalized to 0-1)
      h: 0.48333,
      s: 0.6,
      l: 0.64,
    }
  },
  // Other colors
  {
    ...
  }
]

If this crate is used in some Rust projects, then following function is also available

pigments_pixels(pixels: &Vec<LAB>, k: u8, weight: fn(&LAB) -> f32, max_iter: Option<u16>) -> Vec<(LAB, f32)>

This function can be used when color data is gathered from an image decoded using image-rs.

Arguments
  • pixels reference to a Vector of colors in LAB format.
  • k defines the number of colors to be gathered from the image.
  • weight defines the weight function to use. src/weights.rs file has few implemented weight functions.
  • max_iter defines the maximum iterations that algorithm makes, default is 300
Return

Returns a vector of tuples with colors as LAB and dominance(as percentage) of each color found in the image.

License

Pigmnts is MIT Licensed

Dependencies

~9–12MB
~226K SLoC