Chapter 2

Artificial Evolution and the Parisian Approach. Applications in the Processing of Signals and Images

2.1. Introduction

This chapter aims to present the so-called Parisian approach and one of its applications in the field of signal and image processing. Modeling, particularly in the fields of geometry and physics, has already been covered in Chapter 1. However, this may also come into play in image processing applications. Optimization through artificial evolution also plays a role in efficiently instantiating these models. In certain cases, it might be better to decompose the model into its elementary entities as they are easier to manipulate. Objects that have been manipulated by genetic operators are no longer vectors of parameters that describe a complete model of a scene. The elementary entities, however, take on a similar role. They only come together as a whole when the representative model of the scene is being studied.

2.2. The Parisian approach for evolutionary algorithms

In traditional evolutionary algorithms, the aim is to find the best possible solution for a given problem. However, certain problems (NP-complete, or not) can turn out to be extremely complex, especially if the search space is very large. To conceptualize the problem the evolutionary algorithm is confronted with, it makes sense to look at some examples. Let us assume that, for instance, we want to segment (isolate) the different objects of an interior scene using a standard evolutionary algorithm. ...

Get Optimisation in Signal and Image Processing now with the O’Reilly learning platform.

O’Reilly members experience live online training, plus books, videos, and digital content from nearly 200 publishers.