An evolutionary algorithm is used to search for iterated function systems (
IFS) that can encode black and white images. As the number of maps of the I
FS that encodes an image cannot be known in advance, a variable-length geno
type is used to represent candidate solutions. Accordingly, feasibility con
ditions of the maps are introduced, and special genetic operators that main
tain and control their feasibility are defined, In addition, several simila
rity measures are used to define different fitness functions for experiment
ation. The performance of the proposed methods is tested on a set of binary
images, and experimental results are reported.