Satellite / AI Imagery
Checking…Sign in
SpatialIQ reconciliation v1

Compare field-captured structures with Open Buildings candidates.

Start with one ward, import a small batch from Google Earth Engine, then use the unmatched and out-of-bounds queues to decide where supervisors or GIS staff should investigate next.

Imagery-derivedfootprints carry import freshness, capture date, sensor and resolution (SIM-seeded metadata until the provider exposes it) and detection confidence

Built-form classification only — no occupant or personal data. This screen classifies built form; it contains no occupant counts, names or personal data.

1. Choose ward2. Import footprints3. Review alerts4. Dispatch follow-up
Field follow-up →
Open Footprints
satellite_alt
—
imported building candidates
Matched
join_inner
—
— reconciliation rate
Unmatched
travel_explore
—
needs field verification
Out of Bounds
rule
—
outside ward or EA geometry
Imagery freshness
update
—
no acquisition date on record
Safe import path: use Real, choose a single ward, keep confidence near 0.7, and start with 100-500 candidates. Demo data is only for training the interface.
Source view

SpatialIQ supervisor alerts

Highest-priority wards based on unmatched candidates and geometry issues. Use this as the dispatch list.

Wards with imported candidates that need supervisor or GIS attention.
WardAlertUnmatchedOut boundsDispatch
No SpatialIQ alerts for the selected source.

Ward reconciliation

Shows whether imported building candidates are already represented by field structures in each ward.

Per-ward totals for matched, unmatched, field-check, and out-of-bounds candidates.
WardStateTotalMatchedUnmatchedField checkOut boundsRate
No open footprints ingested yet.

Matched footprints — imagery confirmed by field GPS

Open Buildings candidates that reconciled to a GPS-anchored field-captured structure within the match radius. This is detection accuracy at the structure level — the imagery footprint verified against the enumerator's house GPS. No occupants; buildings are classified and counted, never attributed a population.

Imagery footprint ↔ field structure pairs (reconciled within the match radius).
FootprintEAField structureHouse GPS (field-captured)Δ match
No matched footprints in this view — reconcile candidates in a ward that has field-captured structures.

Unmatched footprint queue

Building candidates that may need field verification because no nearby captured structure was found.

Candidate building points for enumerator or supervisor follow-up.
CandidateLocationEAConfidenceSource
No unmatched footprints in the queue.

Next in the flow: Classification → Building Intelligence →