Awesome
<!-- WARNINGS: * Bits of information here are duplicated in several places: * https://docs.frameright.io/javascript * https://github.com/Frameright/image-display-control-metadata-parser Make sure to keep them in sync. * Make sure all URLs in this document are absolute, and not relative within GitHub, as we are publishing this file to NPM and want URLs to remain valid there. -->
Image Display Control metadata parsing library
➡️ See this document rendered at docs.frameright.io/javascript
An easy way to retrieve Image Display Control metadata out of images. Made with :heart: by Frameright. Power to the pictures!
:sparkles: Live demo
:bulb: GitHub Discussions
NOTE: this is a wrapper around mattiasw/ExifReader and image-size. Many thanks to mattiasw, netroy, and other contributors!
Table of Contents
<!-- toc --> <!-- tocstop -->Overview
The
Image Display Control web component
extends the <img> tag with the ability to accept a list of
image regions, and to zoom in on the best one for the current element size, thus
achieving better results than
object-fit: cover;
a.k.a. middle-cropping. Its syntax looks like:
<img
is="image-display-control"
src="https://images.pexels.com/photos/3625715/pexels-photo-3625715.jpeg"
width="200"
height="200"
data-image-regions='[{
"id": "oneanimal",
"names": ["One animal"],
"shape": "rectangle",
"unit": "relative",
"x": "0.217",
"y": "0.708",
"width": "0.239",
"height": "0.1467"
}, {
"id": "threeanimals",
"names": ["Three animals"],
"shape": "rectangle",
"unit": "relative",
"x": "0.245",
"y": "0.725",
"width": "0.419",
"height": "0.121"
}]'
/>
Typically this list of image regions come from the metadata of the image file
itself, is retrieved by the back-end, and is placed in the front-end's
<img data-image-regions=" attribute.
This is where this library comes into play: it allows your Node.js back-end to easily retrieve this metadata.
NOTES:
Usage
In a Node.js back-end
#!/usr/bin/env node
// ./myscript.mjs
import { promises as fs } from 'fs';
// npm install @frameright/image-display-control-metadata-parser
import { Parser } from '@frameright/image-display-control-metadata-parser';
// Get it from https://iptc.org/std/photometadata/examples/IPTC-PhotometadataRef-Std2021.1.jpg
const buffer = await fs.readFile('IPTC-PhotometadataRef-Std2021.1.jpg');
const parser = new Parser(buffer);
console.log(parser.getIdcMetadata());
This has been validated with JPEG, PNG, and WebP images.
:memo: Tutorial
:scroll: Reference
:wrench: Contributing
:bulb: GitHub Discussions
Directly in a browser
For testing purposes, you can use this library directly in the browser:
<html>
<body onload="documentLoaded()">
<script
type="module"
src="https://cdn.jsdelivr.net/npm/@frameright/image-display-control-metadata-parser@2.0.0/dist/image-display-control-metadata-parser-standalone.min.js"
></script>
<script type="text/javascript">
async function documentLoaded() {
const image = await fetch(
'https://iptc.org/std/photometadata/examples/IPTC-PhotometadataRef-Std2021.1.jpg'
);
const buffer = await image.arrayBuffer();
const parser = new ImageDisplayControl.Parser(buffer);
const regions = parser.getIdcMetadata();
console.log(JSON.stringify(regions, null, 2 /*indent*/));
}
</script>
</body>
</html>
The parsed metadata can then directly be fed to the Image Display Control web component.
This React component is an example of doing that.
:sparkles: Live demo
Image Display Control metadata
Nowadays an image file (e.g. JPEG, PNG) can contain this type of image regions in their metadata according to the IPTC standard. Photographers, or anyone else, can use the Frameright webapp to define and store image regions in the metadata of their pictures.