Smartcrop.js

Smartcrop (GitHub: jwagner/smartcrop.js, License: MIT, npm: smartcrop, bower: smartcrop)

Smartcrop.js is not just another image cropping library. It detects the primary region of an image such as faces. It does this using the following algorithm.

1. Find edges using Laplace method
2. Find regions with a color like skin
3. Find regions high in saturation
4. Generate a set of candidate crops using a sliding window
5. Rank them using a importance function to focus the detail in the center and avoid it in the edges.
6. Output the candidate crop with the highest rank

Using it in a simple example:

SmartCrop.crop(image, {width: 100, height: 100}, result => console.log(result) );
// Which will return
// -> {topCrop: {x: 300, y: 200, height: 200, width: 200}}

When you wanna use it within node.js you would have to do something like this:

import fs from 'fs';
import Canvas from 'canvas';
import SmartCrop from 'smartcrop';

SmartCrop.crop(img, options, result => {
    console.log(JSON.stringify(result, null, 2));
    if(output && options.width && options.height){
        var canvas = new Canvas(options.width, options.height),
            ctx = canvas.getContext('2d'),
            crop = result.topCrop,
            f = fs.createWriteStream(output);
        ctx.patternQuality = 'best';
        ctx.filter = 'best';
        ctx.drawImage(img, crop.x, crop.y, crop.width, crop.height, 0, 0, canvas.width, canvas.height);
        canvas.syncJPEGStream({quality: quality}).pipe(f);
    }
});

Smartcrop.js also offers a command line interface under the npm package name smartcrop-cli. Which based uses node-canvas.

That one you would use like this:

npm i smartcrop-cli -g
smartcrop-cli --width 100 --height 100 photo.jpg square-thumbnail.jpg

Smartcrop-cli does requires node-canvas which requires local libraries to be installed. Refer to the node-canvas wiki installation instructions.