The activities NR are focusing on are:
Segmentation of avalanche structures: The aim of this work is to develop algorithms for segmenting avalanche structures in high resolution optical satellite images. Texture is one of the important characteristics that make it possible to visually discriminate avalanche affected areas from non-affected areas in the image. An avalanche will typically result in a characteristic "rake" pattern oriented in the aspect direction of the terrain. Texture based processing methods that enhances "rake" patterns in the aspect direction are therefore essential locate potential avalanche structures.
Feature extraction and classification: The aim of this work is to extract features from the segmented avalanche structures that may be used to distinguish avalanches from other segmented objects. Typically such features are based on the shape and structure of the segmented objects and contextual information related to the object.
A flow chart of the avalanche detection and mapping system is summarized in the figure below.
A general system for avalanche detection and mapping.