JNetic Textures

by Steven Bergen (srbtutor AT gmail . com)
Supervisor: Brian Ross

JNetic Textures is a system for evolutionary image generation. It uses genetic programming to generate images. A unique feature is its use of both texture formulae and vector graphics together to generate images. Another feature is its use of mathematical models of aesthetics, such as Ralph's Deviation from Normality analyses. Multi-objective evolution is used to reconcile multiple fitness scores. The system can be run interactively or automatically. It is available for download below.

GPTP 2010 presentation (pptx).

Book chapter: "Evolutionary Art Using Summed Multi-Objective Ranks". S. Bergen and B.J. Ross. Genetic Programming Theory and Practice VIII, R. Riolo, T. McConaghy, E. Vladislavleva (eds.), Springer, 2010, pp. 227-244.
Paper images

Gallery

JNetic system copyright (c) 2010 Steve Bergen.
Generated images copyright (c) 2010 Steve Bergen and Brian J. Ross.
Back up: http://www.cosc.brocku.ca/~bross/