A new fuzzy based flood-fill algorithm for 3D NMR brain segmentation

Segmentation of images is one of the most difficult problems in the field of image processing. Especially medical images contain several problems of inter-individual variability and uncrisp data. In the paper a segmentation algorithm, based on region growing methods (often called flood-fill methods), is proposed. The region growing is supported by fuzzy rules to improve segmentation results. Beside the algorithm some results of the segmentation are presented.