We discuss here images with complex structure such as biocrystals, which are very often turn out to be fractals or multifractals. We present 3 types of multifractal spectra, and vector characteristics based on using blanket technic for the surface of grey-level function constructed for a halftone or monochrome image. Such a set of characteristic describes the image structure quite complete. In this work we
apply several different fractal and multifractal methods to analyze images. Our experiments make it obvious that for every class of images at least 2 methods allow obtaining reliable separation of
numerical signs. The algorithms for calculation multifractal characteristics are implemented. For each class of images the
most appropriate signs were recommended.