The assessment or description of the internal morphology of floes plays an important role in the optimization of flocculation processes and the development of high-performance reaction systems. This paper is focused on mathematical algorithms-namely c-means clustering, fuzzy clustering, Fourier transformation and kernel smoothing-to reconstruct the internal cluster structure of floes. In principle, all methods are suitable to interpret the growth history of floes. However, it is shown that at the moment c-means clustering in combination with a data preparation step seems to be the most favorable procedure in that context. A detailed description of the data preparation procedure is not part of this work.