This is a little utility for generating heightmaps from Shuttle Radar Topography Mission records.
They may be somewhat visually meaningful on their own, but mostly they're a convenient way to pack three dimensional data into a format browsers understand for applications like this.
1px = 1/3600 degree longitude or latitude, about 30m, color value is height, encoded as described below
pip install numpy PIL, run as ./hgt2png or add to PATH
Download and create PNG of the SRTM data for N 38 W 97 to N 39 W 96, in 1 big tile containing the point [lon,lat] provided:
hgt2png --download 96.432 38.123
Download and create a PNG of the data within a bounding box:
hgt2png --bbox -96.8 38.2 -96.6 38.5
Use, resize or reuse a file you already downloaded:
hgt2png --file N38W097.hgt --bbox -96.8 38.2 -96.6 38.5
Render a square area only:
hgt2png --square --file N38W097.hgt --bbox -96.8 38.2 -96.6 38.5
Red channel: floor ( height / 256 ) (measurement in meters)
Green channel: height mod 256
Blue channel: centimeters offset from r*256 + b (not used in SRTM,
sort of left here in case some client stuff ever gets written
that could use more precision.)
Alpha channel: Just a switch indicating negative or positive. 255 or
any other value means positive, 21 (0x15 the ASCII code for negative
acknowledge character (tee hee)) means negative.
Elevation equals ( RED * 256 + GREEN + BLUE/100 ) * (ALPHA == 21) * (-1) Unprojected lon lat coordinates can be calculated from the images + file names.
you can call it with --exportable and you'll get a georeferenced tif
that you can use for example in TileMill or QGIS.