ABSTRACT
This chapter discusses a combinatorial optimization method: simulated annealing. Optimization problems which occur in technological applications can quickly become forbiddingly complex and require huge computation times because of the combinatorial explosion of the number of possible solutions. The chapter also discusses lower-level processing to reconstruct the image or to extract characteristic features from it. The task assignment problem is that of determining the most efficient distribution. The idea behind simulated annealing in combinatorial optimization is to numerically simulate a thermal annealing operation. A number of cities are distributed on a map and the traveling salesman must organize a tour which visits each city exactly once before returning to the starting point. Image processing is a set of techniques which improves images and enables them to be interpreted. The connection between magnetic systems at a finite temperature, probabilistic automata, and simulated annealing is clear in the examples of the distribution of logic functions between different integrated circuits, and of image restoration.
