Classification of shape and texture in digitized images has focused the interest on parameters like the fractal dimension. There are several notions of fractal dimensions [6]. The box dimension and some relatives have gained importance in image processing and have proven their usefulness [5].
The following article describes the algorithms which have been implemented in the image processor IDOLON. It performs fractal image analysis on digitized images both from the entire image and from selected regions. The parameter estimates can be taken to supply classification problems e.g. from satellite images.