Snow avalanche detection and classification algorithm for GB-SAR imagery

This paper presents an algorithm for the automatic detection and classification of snow avalanches using GB-SAR imagery. The algorithm has been validated extensively in two different test sites in Sion (CH) and Alagna (I), respectively. Results show an important reduction on the images to manually supervise and minimal false negative rate. The real-time availability of the location and extent of the avalanche events provided by this system has demonstrated to be of high interest to ski resort administrators, allowing a more efficient management of the resources available to prepare the tracks.