Skinning a virtual Ant Image (AI)

There are many way to skin an ant. In blender, gimp, SVG and life there are multiple ways to achieve the same objective. In space there are infinite paths to and from a point. In blender I have chosen to try a new way of applying materials to surfaces generated by a simulated cell growth algorithm. The structure of the creature is created in simulation in a way that is consistent with cell growth and differentiation. The matrix is stripped for several applied processes. The "cells" on the surface are kept as a surface for the representation of form in blender and the muscle structure is connected to actuators. The nerve connections are simulated with a mock neural array and then translated to action sequences.

Because I am a lazy programmer -- I would never sit for hours playing with surfaces to create a single model, when I can script a generic process that takes information from a scene, camera, image, or generated from a putative DNA mechanism.

The advantage in this is that I can make a few changes in the DNA sequence and make vast changes in the final result. It requires an understanding of the effect of changes on the final product. This is true of any complex task, like skinning a model or generating an interesting image. If you cannot grasp the relationship between the methods and product, it is very frustrating.

0 comments:

Contributors

Automated Intelligence

Automated Intelligence
Auftrag der unendlichen LOL katzen