Roya is a small low-level image comparison library.

Inspired by pixelmatch with more bells and whistles.

Installation

using yarn

yarn add roya

using npm

npm i -S roya

API

roya(img1, img2, [, options])

img1, img2 — Image data of type Buffer of the images to compare

options Optional object options with the following properties:

Property	Description	Default
threshold	Matching threshold, ranges from 0 to 1. Smaller values make the comparison more sensitive	0.1
highlightFade	Highlight color intensity will depend on the pixel ditance value	true
highlightColor	The used highlight color, should be an array of [R,G,B] values	[255, 0, 0]
transparent	Whether to keep the original image ot to use a blank convas to highlight the diff	false
overlapse	Whether to use the highlight color or use the changing pixel itself	false
method	Matching and diffing method to be used, should be one of the available methods, rgb, rgbTuned, yiq or yiqTuned	'rgb'

P.S. rgb method is the fastest but it might not be the best for your usecase, try the others method and pick the right one for you.

Example:

img1	img2

options	result
undefined
{ threshold: 0.5 }
{ highlightColor: [255, 255, 0] }
{ highlightFade: false }
{ transparent: true, highlightFade: false }
{ overlapse: true, transparent: true }
{ method: 'rgb', highlightFade: false }
{ method: 'rgbTuned', highlightFade: false }
{ method: 'yiq', highlightFade: false }
{ method: 'yiqTuned', highlightFade: false }

Usage

const fs = require('fs')
const PNG = require('pngjs').PNG
const roya = require('roya')

var img1 = fs
  .createReadStream('img1.png')
  .pipe(new PNG())
  .on('parsed', parsed)

var img2 = fs
  .createReadStream('img2.png')
  .pipe(new PNG())
  .on('parsed', parsed)

let done = 0
function parsed() {
  done++
  if (done < 2) return
  const out = roya(img1.data, img2.data)

  const diff = new PNG({ width: img1.width, height: img1.height })

  diff.data = out.data

  diff
    .pack()
    .pipe(fs.createWriteStream('diff.png'))
    .on('finish', () => console.log('done'))
}

P.S. image dimensions must be equal.