Pick two registered COGs that cover the same ground. VoltGrid samples them onto a shared grid, scores change, classifies severity, and drops hotspots on the map — optionally clipped to a PLS-CADD conductor buffer and written into Transmission findings. Built for vegetation managers and storm desks who already collect pre/post orthos.
POST /api/corridor-damage/analyze
{
beforeCogId, afterCogId,
mode: "spectral"|"height"|"change",
conductorDbId?, bufferMeters: 30,
writeFindings: true,
findingCategory: "vegetation"
}
→ hotspots[] + summary + corridorGeoJson?
GET /api/corridor-damage/capabilities
Watch the full change-detection run over Lisbon: LiDAR height-proxy mode classifies structure and canopy delta across 408 flagged hotspots on the transmission right-of-way.
Transmission Analyzer’s TlVegetation engine computes NESC / NERC FAC-003 clearances. Corridor Damage is the map-driven before/after raster workflow. Use both: clearances for compliance, change detection for patrol prioritization after a storm or grow-in season.
ΔNDVI hotspots ranked minor → critical, ready for crew routing and work orders.
Brightness change mode for cloudy or single-band post-event collects.
DEM vs DSM (or multi-date DSM) flags structure / canopy gain-loss without a new LiDAR stack.
Honest boundary: Phase 1 is windowed change detection over geographic COGs — not eCognition / object-based segmentation, and not a point-cloud classifier. Geographic (EPSG:4326) COGs recommended.