Abstract
Emergence is the process of deriving some new and coherent structures, patterns and properties in a complex system. Emergent phenomena occur due to interactions (non-linear and distributed) between the elements of a system over time. An important aspect concerning the emergent phenomena is that they are observable on a macroscopic level, whereas they are produced by the interaction of the elements of the system on a microscopic level. In this paper, we attempt to grab some emergence and complexity principles in order to apply them for problem solving. As an application, we consider the edge detection problem a key task in image analysis. Problem solving by emergence consists in discovering the local interaction rules, which will be able to produce a global solution to the problem that the system faces. More clearly, it consists in finding the local rules which will have some awaited and adequate global behavior, to solve a given problem. This approach relies on evolving cellular automata using a genetic algorithm. The aim is to find automatically the rules that allow solving the edge detection problem by emergence. For the sake of simplicity and convenience, the proposed method was tested on a set of binary images,. Very promising results have been obtained.