i.segment

i.segment segments SAR-derived imagery within GRASS GIS to detect and delineate landslide-affected areas by identifying changes in radar backscattering for rapid post-disaster assessment.

Key Features:

• Rapid Areal Assessment: Enables rapid assessment of landslide extent over large spatial scales using SAR-derived segmentation.
• Pre-processing and Elaboration: Processes radar data to extract inputs required for segmentation and rapid post-disaster analysis.
• Segmentation Capabilities: Segments radar-derived data by detecting patterns in backscatter contrast to delineate landslide-affected areas.

Scientific Applications:

• Landslide detection and mapping: Detects and delineates landslide-affected regions by analyzing contrast in radar backscattering values before and after events, demonstrated on the Tagari River valley, Papua New Guinea, following the M7.5 earthquake on February 25, 2018.

Methodology:

Processes SAR data to identify changes in radar backscattering values that indicate landslide activity; identifies segmentation patterns that correlate with impacted areas; and performance was validated against ground truth data.