Skip to content

Instantly share code, notes, and snippets.

@celoyd
Last active December 27, 2015 23:49
  • Star 2 You must be signed in to star a gist
  • Fork 0 You must be signed in to fork a gist
Star You must be signed in to star a gist
Save celoyd/7408773 to your computer and use it in GitHub Desktop.

I've rendered out two map layers with pre-typhoon Landsat 8 imagery. This is very low resolution (15 m) but may help fill in major highways and rivers for areas that still lack better coverage. They include clouds because I found that splicing them together can introduce seam artifacts that could be mistaken for roads.

The layer templates are:

(Please forgive inconsistent naming.)

This is public domain data on a staff account. Distribute at will.

The layers have highly exaggerated color and contrast. Muddy water and bare soil including roads will generally appear red-brown to purplish. I experimented with false color, but it doesn’t offer much extra clarity in this case and can be confusing for people unused to it (e.g., roads can look blue and be interpreted as rivers).

If this is useful, I’ll take requests for neighboring areas etc. and render them as time permits. For anyone replicating, I used basically this process but with these convert arguments for the color mix and pansharpening:

convert -compose Colorize band-8-projected.tif rgb-upsampled.tif -channel G -gamma 0.95 -channel R -gamma 1.05 -channel RGB -sigmoidal-contrast 80x16% -composite c.tif 

(Results may vary depending on ImageMagick version. Yes, they change stuff like this in point releases.)

To get in touch: @vruba, charlie@mapbox.com, celoyd@gmail.com.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment