Daniel de Souza Carvalho

Daniel has a MSc in electrical engineering (Mackenzie University, Brazil) and computer engineering (Braz Cubas University, Brazil) and is a partner of A2F, a Brazilian IT company specialist in enterprise systems such as databases, web development, virtualization, data centers, system architecture, business intelligence, migration, interfaces, and others.

He is a Mathematica and NKS enthusiast, author of more then 70 Demonstrations, and a NKS Summer School 2007 alumni.

His interests lie in NKS and cellular automata in particular; and other theoretical and applied computer science subjects like artificial intelligence; evolutionary algorithms; computer vision; parallel computing; neural networks; swarm intelligence; biologically inspired algorithms; databases; and system development, analysis, architecture, and electronics (Arduino projects).

Cellular Automaton Classification by Image Processing Technique

Searching among the universe of possible simple rules of any system such as cellular automata, Turing machines, tag systems, substitution systems, and others is a very challenging issue in terms of sophistication and computational power required. With a slight color (k) or neighborhood (r) addition to simple rules, the number of possible computational spectrum of work grows up. To avoid the need of human visual verification of big sets of rules at the computational universe, high-quality search algorithms would be appropriate to identify complex patterns cases.

Finding an automated classification of CA complexity is a important tool to support NKS and other computer science research.

This project consists of using image processing techniques like set theory, topology, discrete mathematics, and mathematical morphology in order to check its efficiency to classify cellular automata spaces (totalistic r=1 and k=4 in particular) and identify Wolfram's four classes of behavior of each rule.

Rule 935912