extremelyjesse
652
8
2
This is a computational experimental art exploration inspired by the work of Jackson Pollock.
The basic premise is: what if you could take a digital image, turn its pixels into balls of "paint", and splat them onto a canvas? What would the end result look like? What images / genres of images would produce the most interesting results? Is the original work still recognizable from the end product?
I coded this up in Python primarily using the CV2 library, and beyond selecting the input images and modulating certain experimental parameters, did nothing else to modify the program's output. Below are some of my favorite outputs.
Original: Starry Night, Van Gogh
A personal favorite of mine-- something about the diffusion of yellow-tone pixels creating its own star-like pattern is a fun incidental experimental outcome
Original: The Persistence of Memory, Dali
Would Dali love or hate this? Who's to say, but I'm a fan, and I love that the geometry still "melts" in the post-image.
Original: Individual Print from Marilyn Monroe, Warhol
This was the first Warhol I tried out, and I just adore how his flat color palettes produce such distinct shapes.
Original: Collection of six prints from Marilyn Monroe, Warhol
Possibly my favorite piece in this exploration. The strong shapes that also blend together and produce almost brush-like smudge effects, with a splattering of stray contrasting pixels on "top"-- this has been my phone lock screen since its generation.
Original: Eight prints from Campbell's Soup Cans, Warhol
This was an exploration of experimental parameters that I had a lot of fun with: the "shape" of the splatter, the resolution of the image, and the "orientation" of the RGB cube before triggering the splat. There were a lot more intermediate variations along the way, but I feel this set of 8 best demonstrates the range of options.