Robert Spann
I am intrigued by analogs between the compositional rules and color theory principles that artists use and the mathematical/statistical properties of images. For example, equations, just like images, have symmetries. A digital image is a map from the unit square to a set of k colors. Equations with symmetries can be used to produce images with those same symmetries. Currently, I am experimenting with combinations of equations (formed using Discrete Cosine Transforms) that have different symmetries and/or parities. Combining equations with different symmetries can be used to produce images which have structure, but no symmetries.
I start with three maps from the unit square onto the set of integers {1,2,3}. These maps are constructed using Discrete Cosine Transforms. Two of the maps have horizontal symmetry. One has symmetry with respect to a 180 degree rotation. I combine these three maps to obtain a map from the unit square into the set of integers {1,…,27}. I then take this map Mod 4 (the initial numbers of colors in my desired palette). Each integer is assigned a color. The resulting image is digitally manipulated to achieve a desired aspect ratio and further refined based on my own aesthetic judgments.