algorithmic modeling for Rhino

I hadn't done a stitch of work on this until I happened to come up with the title, then I couldn't resist seeing where it led. See below for a much more refined version using a different approach.

Views: 456


You need to be a member of Grasshopper to add comments!

Join Grasshopper

Comment by DanaLadd on January 23, 2017 at 6:53pm

Thanks Ethan, this post is both helpful and inspiring! Cheers xoxoxo

Comment by Ethan Gross on October 28, 2016 at 1:39am

Hi Dana, and thanks for the "wow"!

There are no plugins involved, just the GH voronoi component. In order to get a decent image, the challenge is to assign more voronoi cells where the tonality changes most dramatically and less where is no variation. Since the time you commented, I have come up with what I think is
an even better way, but both the "classic" version and the new involve two passes through the image - and as of now, only works with nudes...
Method 1 - First looks at the mean-square deviation of intensity within a grid placed over the image and assigned points for the voronoi based on that. The more deviation, the more points, the more points, the more voroinoi cells.

Method 2 - Creates a heightfield landscape based on image intensities and then sends random horizontal lines through it, looks the points of intersection and then projects these to the xy plane. The more variation, the more hills, the more hills, the more intersections. The projected points are then used for the voronoi cells. The second method looks better and takes much less time to run than the first (which isn't saying much)

Method 2 Heightfield

Method 2 top - Intersection of lines with heightfield are projected to xy plane and used for voronoi.

Method 2 result

Script is here:

Comment by DanaLadd on October 27, 2016 at 9:24pm

Wow! What plugin is this?





  • Add Photos
  • View All

© 2021   Created by Scott Davidson.   Powered by

Badges  |  Report an Issue  |  Terms of Service