Skip to content

Latest commit

 

History

History
345 lines (271 loc) · 23.7 KB

README.md

File metadata and controls

345 lines (271 loc) · 23.7 KB
Raw

@thi.ng/pixel

npm version npm downloads Mastodon Follow

Note

This is one of 199 standalone projects, maintained as part of the @thi.ng/umbrella monorepo and anti-framework.

🚀 Please help me to work full-time on these projects by sponsoring me on GitHub. Thank you! ❤️

About

Typedarray integer & float pixel buffers w/ customizable formats, blitting, drawing, convolution.

Important

In July 2024 this package was restructured and split-up to extract some features into smaller more focused packages:

  • Buffer creation from HTML image elements or canvas w/ opt resize & format conversion (browser only)
  • 12 packed integer and 6 floating point preset formats (see table below)
  • Palette-based indexed pixel formats
  • Buffer-to-buffer blitting w/ automatic format conversion
  • Buffer-to-canvas blitting (incl. offscreen canvas support)
  • Buffer-to-buffer blending w/ Porter-Duff operators
  • Pre/post-multiply alpha
  • Region / sub-image extraction
  • Single-channel manipulation / extraction / replacement / conversion
  • Accessors for normalized channel value
  • Image sampling & filtered resizing
    • Filters: nearest neighbor, bilinear, bicubic
    • Wrap behaviors: clamp, wrap, repeat
  • Invert image
  • XY coordinate-based pixel & channel-only accessors (w/ optional bounds checking)
  • Declarative custom pixel formats with optimized code generation
  • HTML canvas creation & ImageData utilities

Integer pixel formats

All integer formats use the canvas native ABGR 32bit format as common intermediate for conversions. During conversion to ABGR, channels with sizes smaller than 8 bits will be scaled appropriately to ensure an as full-range and as linear as possible mapping. E.g. a 4 bit channel will be scaled by 255 / 15 = 17.

Format specs can freely control channel layout within current limits:

  • Channel sizes: 1 - 32 bits.
  • Storage: 8, 16 or 32 bits per pixel

Custom formats can be defined via defIntFormat().

Format ID Bits per pixel Description
ALPHA8 8 8 bit channel (alpha only)
GRAY8 8 8 bit single channel (grayscale conv)
GRAY_ALPHA8 16 8 bit single channel (grayscale conv), 8 bit alpha
GRAY16 16 16 bit single channel (grayscale conv)
GRAY_ALPHA16 32 16 bit single channel (grayscale conv), 16 bit alpha
ARGB4444 16 4 channels @ 4 bits each
ARGB1555 16 5 bits each for RGB, 1 bit alpha
RGB565 16 5 bits red, 6 bits green, 5 bits blue
RGB888 32 (24 effective) 3 channels @ 8 bits each
ARGB8888 32 4 channels @ 8 bits each
BGR888 32 (24 effective) 3 channels @ 8 bits each
ABGR8888 32 4 channels @ 8 bits each
  • ALPHA8 is mapped from/to ABGR alpha channel
  • GRAY8/16, GRAY_ALPHA8/16 compute grayscale/luminance when converting from ABGR and in return produce grayscale ABGR
  • In all built-in formats supporting it, the alpha channel always occupies the most-significant bits (up to format size)

Indexed, palette-based pixel formats

Instead of storing colors directly for each pixel, palette-based formats are supported which only store a color index per pixel (e.g. as is done for GIF and/or indexed PNG formats). These formats can be created via the defIndexed() family of functions.

Floating point pixel formats

Strided floating point format presets for use with floatBuffer(). New formats can be defined via defFloatFormat().

Format ID Channel count Description
FLOAT_GRAY 1 Single channel / grayscale
FLOAT_GRAY_ALPHA 2 Grayscale and alpha channel
FLOAT_GRAY_RANGE 1 Grayscale (user defined value range)
FLOAT_NORMAL 3 Normal map (signed values)
FLOAT_RGB 3 Red, Green, Blue
FLOAT_RGBA 4 Red, Green, Blue, Alpha
  • All color channels are unclamped (but can be clamped via buf.clamp()). For conversion to packed int formats assumed to contain normalized data (i.e. [0..1] interval, with exception of FLOAT_NORMAL which uses [-1..1] range)
  • Conversion between float formats is currently unsupported

Filtered image sampling and resizing

Available (and optimized) for both integer & floating point formats, image samplers can be created with the following filters & wrap modes:

Filters

  • "nearest" - nearest neighbor
  • "linear" - bilinear interpolation
  • "cubic" - bicubic interpolation

Wrap mode

  • "clamp" - outside values return 0
  • "wrap" - infinite tiling
  • "repeat" - edge pixels are repeated
import { intBuffer, defSampler, ABGR8888 } from "@thi.ng/pixel";

const src = intBuffer(4, 4, ABGR8888);

// fill w/ random colors
src.forEach((_,i) => 0xff000000 | Math.random() * 0xffffff);

// create bilinear sampler w/ repeated edge pixels
const sampler = defSampler(src, "linear", "repeat");

// sample at fractional positions (even outside image)
sampler(-1.1, 0.5).toString(16)
// 'ff79643a'

// resize image to 1024x256 using bicubic sampling
const img = src.resize(1024, 256, "cubic");
Filter
"nearest" resized image w/ nearest neighbor sampling
"linear" resized image w/ bilinear sampling
"cubic" resized image w/ bicubic sampling

Status

STABLE - used in production

Search or submit any issues for this package

Support packages

Related packages

  • @thi.ng/color - Array-based color types, CSS parsing, conversions, transformations, declarative theme generation, gradients, presets
  • @thi.ng/porter-duff - Porter-Duff operators for packed ints & float-array alpha compositing
  • @thi.ng/rasterize - Headless 2D shape drawing, filling & rasterization for arbitrary targets/purposes (no canvas required)
  • @thi.ng/shader-ast - DSL to define shader code in TypeScript and cross-compile to GLSL, JS and other targets
  • @thi.ng/webgl - WebGL & GLSL abstraction layer

Installation

yarn add @thi.ng/pixel

ESM import:

import * as pix from "@thi.ng/pixel";

Browser ESM import:

<script type="module" src="https://esm.run/@thi.ng/pixel"></script>

JSDelivr documentation

For Node.js REPL:

const pix = await import("@thi.ng/pixel");

Package sizes (brotli'd, pre-treeshake): ESM: 7.35 KB

Dependencies

Note: @thi.ng/api is in most cases a type-only import (not used at runtime)

Usage examples

26 projects in this repo's /examples directory are using this package:

Screenshot Description Live demo Source
Interactive image processing (adaptive threshold) Demo Source
ASCII art raymarching with thi.ng/shader-ast & thi.ng/text-canvas Demo Source
Interactive & reactive image blurhash generator Demo Source
Color palette generation via dominant color extraction from uploaded images Demo Source
2.5D hidden line visualization of digital elevation files (DEM) Demo Source
Barnsley fern IFS fractal renderer Demo Source
Pixel buffer manipulations Demo Source
Matrix-based image color adjustments Demo Source
Showcase of various dithering algorithms Demo Source
Randomized 4-point 2D color gradient image generator Demo Source
Image dithering and remapping using indexed palettes Demo Source
Normal map creation/conversion basics Demo Source
Interactive pixel sorting tool using thi.ng/color & thi.ng/pixel Demo Source
RGB waveform image analysis Demo Source
Image-based Poisson-disk sampling Demo Source
Port-Duff image compositing / alpha blending Demo Source
Steering behavior drawing with alpha-blended shapes Demo Source
Basic usage of the declarative rdom-forms generator Demo Source
Responsive image gallery with tag-based Jaccard similarity ranking Demo Source
2D scenegraph & image map based geometry manipulation Demo Source
WebGL & Canvas2D textured tunnel shader Demo Source
Fork-join worker-based raymarch renderer (JS/CPU only) Demo Source
Textmode image warping w/ 16bit color output Demo Source
Multi-layer vectorization & dithering of bitmap images Demo Source
Visual comparison of biased vs. unbiased normal vectors projected on the surface of a sphere Demo Source
Minimal multi-pass / GPGPU example Demo Source

API

Generated API docs

import * as pix from "@thi.ng/pixel";
import { SRC_OVER_I } from "@thi.ng/porter-duff";
import { pixelCanvas2d } from "@thi.ng/canvas";

import IMG from "../assets/haystack.jpg";
import LOGO from "../assets/logo-64.png";

const [img, logo] = await Promise.all([IMG, LOGO].map((x) => imageFromURL(x)));

// init 16bit int RGB pixel buffer from image (resized to 256x256)
const buf = intBufferFromImage(img, RGB565, 256, 256);

// create grayscale buffer for logo and use Porter-Duff operator to
// composite with main image. Since the logo has transparency, we
// need to premultiply alpha first...
intBufferFromImage(logo, GRAY_ALPHA8).premultiply().blend(SRC_OVER_I, buf, {
    dx: 10,
    dy: 10,
});

// extract sub-image
// (method returns undefined if result region is < 1 pixel)
const region = buf.getRegion(32, 96, 128, 64)!;
// copy region back at new position
region.blit(buf, { dx: 96, dy: 32 });

// or alternatively blit buf into itself
// buf.blit(buf, { dx: 96, dy: 32, sx: 32, sy: 96, w: 128, h: 64 });

// create html canvas
// (returns obj of canvas & 2d context)
const { canvas } = pixelCanvas2d(buf.width, buf.height * 3, document.body);

// write pixel buffer to canvas
buf.blitCanvas(canvas);

// manipulate single color channel
const id = 0;
const ch = buf.getChannel(id).invert();
for (let y = 0; y < ch.height; y += 2) {
    for (let x = (y >> 1) & 1; x < ch.width; x += 2) {
        ch.setAt(x, y, 0xff);
    }
}
// replace original channel
buf.setChannel(id, ch);
// write pixel buffer to new position
buf.blitCanvas(canvas, { y: buf.height });
// create & write grayscale version
buf.as(GRAY8).blitCanvas(canvas, { y: buf.height * 2 });

Authors

If this project contributes to an academic publication, please cite it as:

@misc{thing-pixel,
  title = "@thi.ng/pixel",
  author = "Karsten Schmidt and others",
  note = "https://thi.ng/pixel",
  year = 2019
}

License

© 2019 - 2024 Karsten Schmidt // Apache License 2.0