nodeca/pica

pica - high quality image resize in browser

Resize images in browser without pixelation and reasonably fast. Autoselect the best of available technologies: webworkers, webassembly, createImageBitmap, pure JS.

demo

With pica you can:

• Reduce upload size for large images, saving upload time.
• Save server resources on image processing.
• Generate thumbnails in browser.
• ...

Note. If you need File/Blob resize (from form's file input), consider using image-blob-reduce. It has additional machinery to process orientation, keep EXIF metadata and so on.

Migration from pica v9 to pica v10

• If you targeted IE or other legacy browsers — drop them, only modern browsers are supported now.
• If you used new on the default export (new (require('pica'))() or import Pica from 'pica'; new Pica()) — switch to either the factory call require('pica')() / pica(), or to the named Pica class: import { Pica } from 'pica'; new Pica(). The default export is now a factory function only.
• If you passed createCanvas in options — remove it. To override canvas creation, expose a custom OffscreenCanvas on the global scope instead.
• If you used the positional quality argument (pica.resize(from, to, 3)) — switch to pica.resize(from, to, { filter: 'lanczos3' }). The object form { quality: 3 } also works but is deprecated, prefer filter.
• If you relied on multiple Pica instances sharing the same worker pool — refactor to create a single instance and reuse it. Implicit sharing is gone.

Prior to use

Here is a short list of problems you can face:

• Loading image:
  • Due to JS security restrictions, you can process images from the same domain or local files only. If you load images from a remote domain, use the proper Access-Control-Allow-Origin header.
  • iOS has memory limits for canvas elements, which may cause problems in some cases, more details.
  • If your source data is a jpeg image, it can be rotated. Consider using image-blob-reduce.
• Saving image:
  • Some ancient browsers do not support the canvas.toBlob() method. Use pica.toBlob(), it includes the required shim.
  • For jpeg source, it's a good idea to keep exif data. Consider using image-blob-reduce.
• Quality
  • JS canvas does not support access to info about gamma correction. Bitmaps have 8 bits per channel. That causes some quality loss, because with gamma correction precision could be 12 bits per channel.
  • Precision loss will not be noticeable for ordinary images like kittens, selfies and so on. But we don't recommend this library for resizing professional quality images.

Install

npm install pica

Use

// ESM (default factory)
import pica from 'pica';
const resizer = pica();

// ESM (class)
import { Pica } from 'pica';
const resizer = new Pica();

// CommonJS
const resizer = require('pica')();

// Resize from Canvas/Image to another Canvas
resizer.resize(from, to)
  .then(result => console.log('resize done!'));

// Resize & convert to blob
resizer.resize(from, to)
  .then(result => resizer.toBlob(result, 'image/jpeg', 0.90))
  .then(blob => console.log('resized to canvas & created blob!'));

By default, the main entry (pica) inlines the webworker code. If you want the worker as a separate file (smaller main bundle, easier CSP), use the split build — pica/dist/pica_main.mjs together with pica/dist/pica_worker.js. In that case workerURL is required:

import createPica from 'pica/dist/pica_main.mjs';

const resizer = createPica({
  workerURL: new URL('pica/dist/pica_worker.js', import.meta.url)
});

[!NOTE] For a quick look at dist/ folder contents, see https://unpkg.com/pica@latest/.

API

new Pica(config)

Create resizer instance with given config (optional):

• tile - tile width/height. Images are processed by regions, to restrict peak memory use. Default 1024.
• features - list of features to use. Default is [ 'js', 'wasm', 'ww' ]. Can be [ 'js', 'wasm', 'cib', 'ww' ] or [ 'all' ]. Note, cib is buggy in Chrome and does not support the default mks2013 filter.
• idle - cache timeout, ms. Creating webworkers is not fast, so this option allows reusing them effectively. Default 2000.
• workerURL - URL for pica_worker.js when using split builds (pica_main.js / pica_main.mjs) with ww enabled. Full builds (pica.js / pica.mjs) include worker code and do not need this option.
• concurrency - max webworkers pool size. Default is autodetected CPU count, but not more than 4.

Important! Latest browsers may support resize via createImageBitmap. This feature is supported (cib) but disabled by default and not recommended for use. So:

• createImageBitmap() is used for non-blocking image decode (when available, without downscale).
• Its resize feature is blocked in the default pica config. Enable it only at your own risk. The result with cib enabled will depend on your browser. The result without cib will be predictable and good.

.resize(from, to, options) -> Promise

Resize image from one canvas (or image) to another. Sizes are taken from source and destination objects.

• from - source, can be Canvas, Image or ImageBitmap.
• to - destination canvas, its size is supposed to be non-zero.
• options - object:
  • quality (deprecated, use .filter instead) - 0..3.
  • filter - filter name (Default - mks2013). See resize_filter_info.ts for details. mks2013 does both resize and sharpening, it's optimal and not recommended to change.
  • unsharpAmount - >=0. Default = 0 (off). Usually value between 100 to 200 is good. Note, mks2013 filter already does optimal sharpening.
  • unsharpRadius - 0.5..2.0. By default it's not set. Radius of Gaussian blur. If it is less than 0.5, Unsharp Mask is off. Big values are clamped to 2.0.
  • unsharpThreshold - 0..255. Default = 0. Threshold for applying unsharp mask.
  • cancelToken - Promise instance. If defined, current operation will be terminated on rejection.

Result is Promise, resolved with to on success.

(!) If you need to process multiple images, do it sequentially to optimize CPU & memory use. Pica already knows how to use multiple cores (if browser allows). Create a single Pica instance and reuse it across calls.

.toBlob(canvas, mimeType [, quality]) -> Promise

Convenience method, similar to canvas.toBlob(), but with promise interface & polyfill for old browsers.

.resizeBuffer(options) -> Promise

Supplementary method, not recommended for direct use. Resize Uint8Array with raw RGBA bitmap (don't confuse with jpeg / png / ... binaries). It does not use tiles & webworkers. Left for special cases when you really need to process raw binary data (for example, if you decode jpeg files "manually").

• options:
  • src - Uint8Array with source data.
  • width - src image width.
  • height - src image height.
  • toWidth - output width, >=0, in pixels.
  • toHeight - output height, >=0, in pixels.
  • quality (deprecated, use .filter instead) - 0..3.
  • filter - filter name (Default - mks2013). See resize_filter_info.ts for details. mks2013 does both resize and sharpening, it's optimal and not recommended to change.
  • unsharpAmount - >=0. Default = 0 (off). Usually value between 100 to 200 is good. Note, mks2013 filter already does optimal sharpening.
  • unsharpRadius - 0.5..2.0. Radius of Gaussian blur. If it is less than 0.5, Unsharp Mask is off. Big values are clamped to 2.0.
  • unsharpThreshold - 0..255. Default = 0. Threshold for applying unsharp mask.
  • dest - Optional. Output buffer to write data, if you don't wish pica to create a new one.

Result is Promise, resolved with resized rgba buffer.

What is "quality"

quality is a legacy preset still accepted for backwards compatibility, but deprecated — prefer the filter option. Mapping:

• 0 - Box filter, window 0.5px (filter: 'box')
• 1 - Hamming filter, window 1.0px (filter: 'hamming')
• 2 - Lanczos filter, window 2.0px (filter: 'lanczos2')
• 3 - Lanczos filter, window 3.0px (filter: 'lanczos3')

In the real world you will never need to change the default (mks2013) filter. All these variations were implemented to better understand resize math :)

Unsharp mask

After scale down, an image can look a bit blurred. It's a good idea to sharpen it a bit. Pica has a built-in "unsharp mask" filter (off by default). Set unsharpAmount to a positive number to activate the filter.

Filter parameters are similar to ones from Photoshop. We recommend starting with unsharpAmount = 160, unsharpRadius = 0.6 and unsharpThreshold = 1. There is a correspondence between UnsharpMask parameters in popular graphics software.

Browser support

We didn't have time to test all possible combinations, but in general:

• Top level API should work in all browsers, supporting canvas and typed arrays.
• Webworkers, WebAssembly and createImageBitmap are not required, but they will be used if available.
• If you plan to use only pure math core, then typed arrays support will be enough.

Note. Browsers are the main target platform. For Node.js we strongly recommend sharp — it is faster and produces better quality. If you really need pica in Node.js anyway, it can run in a limited mode (no WebWorkers) by exposing an external canvas library as the global OffscreenCanvas. This is not recommended.

import { Canvas } from '@napi-rs/canvas'; // or any other canvas library
import pica from 'pica';

global.OffscreenCanvas = Canvas;

const resizer = pica(); // WebWorkers will not be used

References

You can find these links useful:

• discussions on stackoverflow: 1, 2, 3.
• chromium skia sources: image_operations.cc, convolver.cc.

pica for enterprise

Available as part of the Tidelift Subscription.

The maintainers of pica and thousands of other packages are working with Tidelift to deliver commercial support and maintenance for the open source packages you use to build your applications. Save time, reduce risk, and improve code health, while paying the maintainers of the exact packages you use. Learn more.