imagemagick-compare

ImageMagick's Compare binding for Node.

Features

• Native bindings to the C/C++ Magick++ library.
• Supports promises, async and sync.
• Supports all metrics.
• Full typescript support.

Prerequisites

You will need the following dependencies installed on your machine. Currently only linux is supported.

• node-gyp dependencies (python3, pkgconfig, make, g++)
• imagemagick
• imagemagick-dev

Installation

npm i imagemagick-compare

Usage

Version

Returns the version of the installed Magick++ library.

import { version } from 'imagemagick-compare';

console.log(version());

Compare

Compares two images using Magick++ and returns the result as a number.

• originalImage: Image buffer of the original image.
• compareWith: Image buffer of the image to compare with.
• options: Object with metric options. More details below.

Promise usage

import * as fs from 'fs';
import { compare } from 'imagemagick-compare';

const original = fs.readFileSync(__dirname + '/assets/Lenna_orig.png');
const compareWith = fs.readFileSync(__dirname + '/assets/Lenna_diff.png');
const options = { metric: 'SSIM', radius: 2 };

compare(originalImage, compareWith, options).then(res => console.log(res));

Async usage

import { compare } from 'imagemagick-compare';

// ...

compare(originalImage, compareWith, options, (err, res) => {
    if (typeof err !== 'undefined') {
        throw err;
    }

    console.log(res);
});

Sync usage

import { compareSync } from 'imagemagick-compare';

// ...

console.log(compareSync(originalImage, compareWith, options));

Compare options

• metric: Defaults to SSIM. List of supported metrics below.
• radius: Will only be used for SSIM and DSSIM metric.
• sigma: Will only be used for SSIM and DSSIM metric.
• k1: Will only be used for SSIM and DSSIM metric.
• k2: Will only be used for SSIM and DSSIM metric.

Supported metrics

• AE: absolute error count, number of different pixels (-fuzz affected)
• DSSIM: structural dissimilarity index
• FUZZ: mean color distance
• MAE: mean absolute error (normalized), average channel error distance
• MEPP: mean error per pixel (normalized mean error, normalized peak error)
• MSE: mean error squared, average of the channel error squared
• NCC: normalized cross correlation
• PAE: peak absolute (normalized peak absolute)
• PHASH: perceptual hash for the sRGB and HCLp colorspaces
• PSNR: peak signal to noise ratio
• RMSE: root mean squared (normalized root mean squared)
• SSIM: structural similarity index