Level-set based extraction and tracking of meteorological objects in satellite images

The level set technique is a shape extraction procedure that works by expanding a starting point or curve and aggregating new points, according to geometric, mathematical, and statistic criteria. This technique has proved to be effective for the extraction of significant objects from various types of images including meteorological images. Using the same technique we developed a system to extract and track meteorological structures of interest, from satellite image sequences. Modifications at various levels have been made on the original algorithm in order to fit it to this particular application. The authors explain what are the distinguishing features of the developed algorithm and show some interesting results