It is known that Iterated Function Systems can lead to image compression for most natural images. There is, however, little literature available concerning practical methods. We lay mathematical set-ups for some known algorithms, work with convergence problems, and present an algorithm to quarantee convergence.