Image analysis is increasingly used for monitoring the evolution of crystallization. The aim of these techniques is to track crystal growth through the quantification of shape changes. However, images for samples taken directly from the crystallization slurry show that crystals are not isolated, but display a complex pattern with an irregular packing of small and large particles. This work explores the application of fractal analysis methods for studying the patterns showed by these images. To this end, the detrended fluctuation analysis (DFA) is used, which provides a scaling exponent index to quantify the complexity degree of an irregular two-dimensional pattern. The results show that the scaling exponent and the mass fractal dimension of the image provides insights in the evolution of the crystallization process. (C) 2010 Elsevier B.V. All rights reserved.