Genetic Programming
Merging biology and programming can lead to some very interesting developments: today I read an article found on Slashdot about the use of genetic programming to produce an image of the Mona Lisa.
Using just 50 semi-transparent polygons a very good version of the painting was produced using a process of random alteration and comparison to the original image. If the new random polygon alterations were closer in similarity to the old alterations then the new set of alterations was saved.
The post on the author’s weblog goes into a little more detail and shows a selection of images to illustrate the process.
Tags: Art, Biology, Evolution, Mona Lisa, Programming
Biology, programming and art!
Funnily enough I used to spend some of my teenage years making some fairly crude evolution sims in Basic. One of these days I’ll have to pick up a proper language and start writing some more. I’d really like to see something that is much more dynamic, rather than working towards a set goal.
I was a rather inspiring creation and similarly to you I’d quite like to explore evolution simulations. I’ve read about evolution simulations before but found this one interesting because it has other potential uses like computationally cheap image production and image compression.
I don’t see that writing a mutli-purpose code for an ‘evolution’ like this program used would be particularly useful. The author says that the generation took three hours and I presume it would take longer if you used code that wasn’t written specifically for generating image production evolution. I was trying to think of other processes which might benefit from a similar program in order to consider if a general evolution program would be effective or useful. What others can you think of?