Parcel-Sense uses computer vision and parcel-aware spatial logic to detect structures directly from aerial imagery. Surface buildings that exist on the ground but are missing, misattributed, or incorrectly recorded in parcel data.
Developed by a Cook County IL assessment inspector focused on parcel-scale reconciliation workflows.
Jurisdictions can have parcels where structures are missing from records, attributed to the wrong lot, or carried forward after the building is gone. Manual review is expensive. Broad imagery scans without parcel context create noise. Parcel-Sense is built to narrow review time onto the parcels that actually matter.
Buildings can exist on the ground but fail to cleanly appear in parcel records, especially when legacy data, workflow gaps, or attribution issues stack over time.
Raw detection alone is not enough. A structure that appears on one parcel may already be correctly recorded on an adjacent lot. Without boundary-aware logic, reviewers waste time on explainable detections.
Datasets can hold onto old assumptions. Demolitions, reconfigurations, and changes on the ground can leave records lagging behind what imagery now shows.
Parcel-Sense is centered on one job: determine whether structures are present on the ground, and where those structures truly belong in parcel context. The system is geared toward practical scanning, triage, and follow-up.
Detect primary and accessory structures directly from aerial imagery instead of relying exclusively on existing records, permits, or manual discovery.
Align detections to parcel geometry so structures can be interpreted in the correct cadastral context rather than as isolated image events.
Check neighboring parcels to explain away edge cases, multi-parcel layouts, and boundary conflicts before they reach a human reviewer.
Support review not only for parcels that appear improved, but also for parcels where the imagery suggests recorded structures may no longer be present.
Turn model signals into structured parcel-level output that can be screened, filtered, and prioritized by field teams, analysts, or internal review staff.
Focus attention on parcels with the clearest divergence between imagery and recorded data so staff time goes where it has the highest value.
Send the parcel list or area to scan. A free 1,000 parcel run is available for qualified first-time evaluations.
Imagery is processed to identify structures, evaluate spatial fit, and screen for parcels where recorded conditions may not align.
Neighboring parcels and parcel boundaries are used to flag explainable detections and improve audit quality.
Receive a structured output ready to analyze on likely mismatches, field screening, or follow-up analysis.
Parcel-Sense uses spatial adjacency logic to flag possible explanations for detections, ensuring reviewers spend more time on meaningful data anomalies.
Region(s) of interest are identified by a CNN on large parcels, enabling targeted zoom-in for more accurate cross-model validation.
A broad image detector can show something interesting. A parcel-aware system can show something operationally useful. That distinction matters when reviewers are deciding where to spend their time.
It scans parcel imagery to determine whether structures appear present on the ground and flags parcels where imagery and recorded data may not align.
The price is 15 cents per parcel with a minimum run charge of $2,750. Qualified first-time evaluations may request a free 1,000 parcel scan.
No. The system is designed around parcel-aware review logic so detections can be interpreted in actual cadastral context instead of as isolated image events.
It can help identify parcels where structures appear to exist on the ground but are missing, misattributed, or otherwise not aligned with recorded parcel data.
If you want to see how Parcel-Sense fits your workflow, start with a free 1,000 parcel evaluation.