Sequential pattern mining of land cover dynamics based on time-series remote sensing images

Remote sensing images constitute a new type of multimedia data well suited to land cover change detection tasks, as they can repetitively provide information about the land surface and its changes over large and inaccessible areas. With plans for more missions and higher resolution earth observation systems, the challenge is increasingly going to be the efficient usability of the millions of collected images, especially the decades of remote sensing image time series, to describe land cover and/or scene evolution and dynamics. In contrast to traditional land cover change measures using pair-wise comparisons that emphasize the compositional or configurational changes between dates, this research focuses on the analysis of the temporal sequence of land cover dynamics, which refers to the succession of land cover types for a given area over more than two observational periods. The expected novel significance of this study is the generalization of the application of the sequential pattern mining method for capturing the spatial variability of landscape patterns and their trajectories of change to reveal information regarding process regularities with satellite imagery. Experimental results showed that this approach not only quantifies land cover changes in terms of the percentage area affected and maps the spatial distribution of these land cover changes but also reveals possibly interesting or useful information regarding the trajectories of change. This method is a valuable complement to existing bi-temporal change detection methods.