Automated image analysis has emerged as a useful tool for quality evaluation and inspection of food processes and products. Image analysis techniques are aimed to the extraction of features for quantifying texture, shapes and distributions of irregular geometries recasted on a microscopy image. The monitoring of crystal growth evolution in traditional industrial processes commonly relies on the visual expertise of long-term trained operators, which limits seriously the automated operation of the process. The objective of this study was to investigate the potential usefulness of fractal metrics; namely Fourier analysis fractal dimension and lacunarity using images, as quantitative descriptors of crystallization evolution. To our knowledge this is the first reported use of lacunarity for the characterization of images of crystallization images from direct samples of crystallization slurries. Fractal dimension and lacunarity increase with the crystallization time. Increased fractal dimension was related to the formation of large clusters in the image, and was taken as an indicative of the amount of formed crystals. On the other hand, lacunarity is an index of non-uniformity of particles on the image, such that lacunarity can be considered as an indicator of the crystal shape and size diversity. In an overall sense, the results showed that fractal analysis can be incorporated as a complementary tool for monitoring the evolution of cane sugar crystallization process. (C) 2010 Elsevier Ltd. All rights reserved.