Research Project
4DEMON: 4D Near Real-Time Environmental Monitoring
Please find our news via those links:
Motivation
Our physical environment underlies permanent changes in space and time with strongly varying triggers, frequencies, magnitudes and also consequences to humans. Monitoring of Earth surface processes (e.g. landslides) and the assessment of environmental properties (e.g. agricultural plant conditions) is crucial to improve our scientific understanding of complex human-environmental interactions and helps us to respond by adaptation or mitigation.
The last decade has witnessed extensive application of 3D environmental monitoring with the LiDAR technology, also referred to as laser scanning. Although a multitude of automatic methods were developed to extract environmental parameters from LiDAR point clouds, only little research has focused on highly multitemporal LiDAR monitoring (4D-LiDAR). Large potential of applying 4D-LiDAR is given for landscape objects with high and varying rates of change (e.g. plant growth), and also for processes with sudden unpredictable changes (e.g. natural hazards).
Main Objective
In this project we (re)assess the scientific concepts and data models for big 4D LiDAR data. In our core concept, a single LiDAR point is treated as an observation in space and time, and the measurements are not independent of each other in space and time. Further, based on two real-world use cases we will develop new algorithms for surface parameter derivation (agricultural crops) and change detection (landslides) making use of the full history
contained in the 4D point cloud time series. We will evaluate our novel methods with respect to near real-time analysis capability (in between of two epochs), making use of the entire big point cloud archive collected during permanent long-term terrestrial LiDAR monitoring.