A hierarchical data fusion framework for vegetation classification from multisource remotely sensed imagery

This paper presents a methodological framework for a hierarchical data fusion system for vegetation classification using multisensor and multitemporal satellite imagery. The uniqueness of the approach is that the overall structure of the fusion system is built upon a hierarchy of remotely sensible attributes of vegetation canopy. This approach also produces classified products that are comprised of a series of important and direct terrestrial variables for ecological modeling with rigorous capabilities across spatial and temporal scales. The framework is mainly consisted of two components: automated image registration and hierarchical model for multisource data fusion.