Tiger pattern algorithm?

Creating a giraffe pattern generator was straight forward (GitHub). Nature doubtlessly designed the fur of these animals with a Voronoi pattern in mind.

Now what about a tiger pattern, with wavy parallel stripes, blobbing together? How could that be best modeled?

Replies to This Discussion

Permalink Reply by Andrew Heumann on February 25, 2017 at 3:22pm

classically I think most "stripe" patterns in nature are generally thought to arise from some kind of activator/inhibitor reaction, described famously by alan turing in "the chemical basis of morphogenesis."

https://phys.org/news/2012-02-alan-turing-1950s-tiger-stripe.html

There are algorithmic models of this kind of limited-diffusion reaction, the grey-scott model being the best known. if you search this forum and the web for "reaction diffusion" you will probably find some resources.

Permalink Reply by Felix E. Klee on February 25, 2017 at 3:40pm

Thanks for the keyword, Andrew!

Indeed there are a few posts here in the forum about R/D. Unfortunately, it seems to be a pixel based algorithm, and so maybe GH is the wrong tool for creating a tiger pattern.

Permalink Reply by Laurent DELRIEU on February 25, 2017 at 3:52pm

You will find some mesh based R/D. I did one

http://www.grasshopper3d.com/forum/topics/reaction-diffusion-on-tri...

you could put directionality to the pattern but it will be not like tiger pattern !

Permalink Reply by Felix E. Klee on February 26, 2017 at 1:24am

Thanks, Laurent, that could be something to play with! Note, that for a start, I only need a 2D pattern.

Permalink Reply by Felix E. Klee on March 5, 2017 at 11:15am

See attachment. Played around a little bit, but the result is far from satisfactory. The approach starts with Delaunay triangulation. Directionality is achieved by removing lines oriented more horizontally. Then there is some smoothing and deformation going on, with the idea of making the stripes fade towards the top.