Just for Fun: Pixel Cuboids

SethF

Just for funsies, I made a model using cuboids as pixels in a picture, using featurescript to assign the color to each 😂😂

Here's the link.

EvanReese

really cool! How'd you preprocess your picture data? I wish Featurescript could do it internally.

john_mcclary

Oh shit. That's awesome. You're going to be at the next user group meeting right? 

SethF

@Evan_Reese I cheated 😂😂. I used an online extractor to pull the data off of an image I had saved at a very low resolution. I thought it'd be cool to do it internally too, but a) didn't know where to begin/if it was possible, and b) wanted to start small by just applying the data to different cubes and see how Onshape handled 2000some (I guess 2772 ) pixels. Very simple, and a very fun start!

@john_mcclary Thanks! I'm gonna try, but it'd have to be while I'm on the road, so we'll see..

xTimRice

Reminds me of the “Polaroid” feature I wrote a while ago  



https://forum.onshape.com/discussion/16650/new-award-winning-custom-feature

bryan_lagrange

Come to Tuesday meeting @SethF and join like minds who push Onshape in all directions..... maybe even @xTimRice will show up.

EvanReese

@SethF
what online tool did you use, if you don't mind? I've been curious to mess with image data. It could be a cool way to adjust the offset of things. You could, for example, use it to create a grid of points at heights based on image values, and create a Fill surface through them to create a 3D printable lithophane.  Or just keep it pixelated like you have but make them different heights for a more stylized look. It would be a fun one to code up!

SethF

@Evan_Reese Yeah! I was thinking of using it for a grid of points for converting relief images (this kinda thing) into a solid model. Obviously already better tools for doing that out there, but it’d be fun to do with this approach. I like the pixelated lithophane idea! That could look super cool. Back to the grid of points; you could probably get a pretty smooth surface for a lithophane cause for one thing you wouldn’t generate nearly 3,000 cubes for a 42x65 pixel photo 😂

I'm pretty sure this is the one I used for extracting the color data. I may or may not have used python to assign the coordinates to the data, just because what little programming experience I do have is mostly in python.. But I do hope to at least write that part in the featurescript soon, ha.

SethF

Here's a just-for-fun variation, with each pixel broken up into it's R, G, and B.

SethF

@xTimRice That is still SO cool! Me and my coworker still geek out about that from time to time!