Raster Blaster

Blast rasters to a canvas with speed and style.

Given multi-band raster data, for example satellite imagery, this module renders them to a canvas, allowing arbitrary mappings of raster bands to canvas color channels. This makes it easy and fast to make contrast adjustments, calculate vegetation index like NDVI and many other common operations.

Tiny example of Raster Blaster in action

Under the hood, uses WebGL when available for high performance, falling back to standard canvas operations if needed.

Note Work in progress or rather proof of concept, probably not suitable for production at this point.

Example

import { Pipeline, WebGlRenderer, PipelineSteps } from 'RasterBlaster'

// A pipeline is a series of functions that are applied to
// the raster data before it is rendered to the canvas
const pipeline = new Pipeline([
  // Convert bands to grayscale using a formula;
  // $r is the incoming "r" band
  new PipelineSteps.GrayScale('$r+$g-$b'),
  // Apply smoothstep to each channel (r, g, b, a)
  new PipelineSteps.SmoothstepContrast(0.2, 0.8),
  // Take the first channel and map its value to rgb values
  // using a named colormap
  new PipelineSteps.ColorMap('RdYlGn'),
  // Set one or more channels directly from raster bands
  new PipelineSteps.BandsToChannels({ a: 'a' })
],
{
  // Map (arbitrary) band names to their indices
  bands: 'rgba',
  dataType: 'Uint8'
})

// A renderer can render raster data to a canvas using a pipeline
// By default, the renderer renders to a 256x256 pixel canvas
const renderer = new WebGlRenderer()
const canvas = document.createElement('canvas')
canvas.width = canvas.height = 256
document.body.appendChild(canvas)

renderer.render(
  canvas,
  pipeline, 
  /* Function that returns a promise that resolves to an array of typed arrays, one for each band */
})