Extract4D
Generalized Extraction of Surface Dynamics from Multimodal 4D Point Clouds for Topographic Monitoring of Earth Surface Processes and their Interactions
Project Description:
Topographic observation of landscape dynamics provides valuable insights into Earth-shaping processes and environmental drivers. As surface dynamics occur on various spatiotemporal scales, better monitoring and characterization of these activities will improve our understanding of environmental processes and human-environment interactions. This information can enhance environmental models of natural processes and human impacts.
Near-continuous 3D observation of local landscapes (i.e. the acquisition of 4D point clouds) significantly enhances the monitoring of surface dynamics. Current time series-based change analysis can effectively extract information about surface activities across various spatiotemporal scales but challenges remain in the transferability and generalization between different data sources (laser scanning and photogrammetry) and different use cases. There is an urgent need to adapt and extend time series-based methods of change analysis in 4D point clouds with adaptive, irregular temporal sampling (from sub-daily to annual) and multimodal sources (e.g., terrestrial and UAV platforms).
This project aims to develop data-driven methods for change analysis using multimodal 4D point clouds with adaptive temporal sampling, enabling automatic extraction of surface activities across different data sources and geographical environments. Such advances will enhance topographic monitoring and support a wide range of environmental research applications.
The developed methods will be applied to several case studies within the project, including coastal dynamics in the Netherlands (in collaboration with Geoscience & Remote Sensing und Coastal Engineering at TU Delft), soil surface processes in agriculture (in collaboration with TU Dresden), and salt marsh dynamics (in collaboration with Geosciences Rennes).