How homssalomssa/information-fusion-suite works
Possibility Fusion is an experimental QGIS Processing plugin that combines Sentinel-2 optical and Sentinel-1 radar change evidence into burned-area possibility maps. The user supplies pre/post rasters, acquisition dates, optional masks and scene parameters; the registered Processing algorithm prepares aligned inputs, derives sensor evidence, applies reliability and temporal handling, fuses evidence, and writes class and diagnostic rasters. The graph preserves documented optional masks and outputs. Algorithm-to-module links are supported by the source index; unsampled algorithm internals limit finer-grained call-flow claims.
Main entry points: fuse_burned_area.py Burned-area algorithm · possibility.py Possibility fusion · plugin.py Plugin lifecycle · provider.py Processing provider
Components 14
QGIS entry
- Plugin lifecycle
possibility_fusion/plugin.py - Processing provider
possibility_fusion/provider.py - Burned-area algorithm
possibility_fusion/algorithms/fuse_burned_area.py
Input preparation
- Acquisition metadata
possibility_fusion/core/acquisitions.py - S2 calibration
possibility_fusion/io/sentinel.py - Raster alignment and IO
possibility_fusion/io/raster.py - Mask and SAR prep
possibility_fusion/core/preprocessing.py - Input validation
possibility_fusion/core/validation.py
Evidence modeling
- Optical dNBR
possibility_fusion/core/indices.py - Sensor memberships
possibility_fusion/core/memberships.py - Temporal reliability
possibility_fusion/core/temporal.py
Fusion outputs
- Possibility fusion
possibility_fusion/core/possibility.py - Map and diagnostics
possibility_fusion/algorithms/fuse_burned_area.py
Other components
- QGIS user
Connections 17
- QGIS user → Plugin lifecycle: loads plugin
- Plugin lifecycle → Processing provider: registers provider
possibility_fusion/plugin.py - Processing provider → Burned-area algorithm: registers algorithm
possibility_fusion/provider.py - QGIS user → Burned-area algorithm: runs processing
README.md - Burned-area algorithm → Acquisition metadata: uses date detection
possibility_fusion/algorithms/fuse_burned_area.py - Burned-area algorithm → S2 calibration: uses calibration
possibility_fusion/algorithms/fuse_burned_area.py - Burned-area algorithm → Raster alignment and IO: uses raster IO
possibility_fusion/algorithms/fuse_burned_area.py - Burned-area algorithm → Mask and SAR prep: uses preprocessing
possibility_fusion/algorithms/fuse_burned_area.py - Burned-area algorithm → Input validation: validates inputs
- Burned-area algorithm → Optical dNBR: derives optical change
possibility_fusion/algorithms/fuse_burned_area.py - Burned-area algorithm → Sensor memberships: forms sensor evidence
possibility_fusion/algorithms/fuse_burned_area.py - Burned-area algorithm → Temporal reliability: applies temporal discount
possibility_fusion/algorithms/fuse_burned_area.py - Burned-area algorithm → Possibility fusion: fuses evidence
possibility_fusion/algorithms/fuse_burned_area.py - Sensor memberships → Possibility fusion: normalizes memberships
possibility_fusion/core/memberships.py - Temporal reliability → Possibility fusion: discounts distributions
possibility_fusion/core/temporal.py - Burned-area algorithm → Map and diagnostics: writes result rasters
possibility_fusion/algorithms/fuse_burned_area.py - Map and diagnostics → QGIS user: returns maps
README.md