Stipple of a hillshade of Tahoma.
The point placing algorithm is freely adapted from @BartWronski’s version of Ahmed, Ren, and Wonka’s stippling: https://bartwronski.com/2022/08/31/progressive-image-stippling-and-greedy-blue-noise-importance-sampling/
The data is from the WA lidar portal: https://lidarportal.dnr.wa.gov/
The hillshade is a mean of 1,000 angles from which the sun actually illuminates Tahoma over the year. (None of that top-left stuff for me, thanks.)
@hans Yes! I also wonder about adjusting it with simulated annealing. I’m sure someone’s already tried that in the literature, but why not try again?
@vruba this is lovely. Makes me want to model the points as little spheres (on a plane? Floating according to height? IDK, it just invites play.)