Using fuzzy mathematical morphology for locating tropical cyclones in satellite imagery

Fuzzy mathematical morphology is an important image analysis tool, especially so when processing images have a certain degree of ambiguity or fuzziness. As the shapes and features of the atmospheric systems observed in remote sensed satellite imagery are inherently ambiguous, fuzzy mathematical morphology seems an appropriate tool for feature extraction and segmentation. In this paper, an algorithm presented for extraction of tropical cyclones from satellite imagery. The segmentation technique is based on extracting the distinguishable features in a tropical cyclone and then making a decision based on the nature of extracted features.