Public & DefenseDigital Twin & Signage Inventory
Kent İkizi
Builds a street-level digital twin of the city from 360° panoramas and camera poses. On that twin, AI finds shop signs, reads them, merges views of the same sign taken from different points and, where the geometry allows, measures its size. The municipality gets its advertising inventory on a map, as a list and as CSV/GeoJSON.
Highlights
- A street-level digital twin from 360° panoramas and camera poses
- Sign detection, segmentation and text reading (OCR) that keeps Turkish characters
- Views from different panoramas merged into one physical sign
- Size and m² from multiple views; a sign that cannot be measured never counts as 0 m², and the reason is recorded
- Human-approved review, inventory map, CSV and GeoJSON export
- Runs on the organisation's own GPU machine; imagery is never sent out
- Measurement validation against LiDAR point clouds
Public & Defense
Try it
A demo running on synthetic data. It holds no real customer data and works fully by keyboard.
Have a similar project?
Bring your idea. We'll scope the right system with you and deliver it.