Innovative automatic designer using a texture-based evolutionary art system

Designing a system for artistic purposes is a challenging task. As judging beauty is a subjective task, most systems rely on Interactive Evolutionary Computation, which the user takes the tedious task to make evaluation decisions. In this paper, we introduce a texture-based evolutionary art algorithm, as an automatic designer system, using genetic algorithm. We mainly focus on two important parts of Genetic Algorithms, initialization and evaluation of individuals in population. We introduced a new method for initializing the population using mathematical functions, including trigonometry formulas that are periodic and create artificial textures. Also, we introduced a new method based on the energy feature to evaluate the population and manage the process of selecting and termination condition of Genetic Algorithm. Considering our proposed automatic termination condition, we removed user fatigue. The proposed initialization method causes a faster convergence, while the resulted images are acceptable by aestheticians.