A robust active contour model driven by fuzzy c-means energy for fast image segmentation.

In this paper, we propose a robust region-based active contour model driven by fuzzy c-means energy that draws upon the clustering intensity information for fast image segmentation. The main idea of fuzzy c-means energy is to quickly compute the two types of cluster center functions for all points in image domain by fuzzy c-means algorithm locally with a proper preprocessing procedure before the curve starts to evolve. The time-consuming local fitting functions in traditional models are substituted with these two functions. Furthermore, a sign function and a Gaussian filtering function are utilized to replace the penalty term and the length term in most models, respectively. Experiments on several synthetic and real images have proved that the proposed model can segment images with intensity inhomogeneity efficiently and precisely. Moreover, the proposed model has a good robustness on initial contour, parameters and different kinds of noise. • We define the cluster center functions and fuzzy c-means energy. • A novel active contour model driven by fuzzy c-means energy is proposed. • An improvement of the penalty and length terms is utilized. • The proposed model shows better segmentation accuracy and robustness. [ABSTRACT FROM AUTHOR]