Parking Sign Detection & Digital Twin

An end-to-end AI pipeline that turns dashcam footage into a structured, map-ready inventory of every parking sign on your network — fully compliant with VicRoads TEM Vol 2 Part 211 and AS1742.11.

Why councils need a parking sign digital twin

Local councils across Australia manage thousands of parking signs. Keeping an accurate, up-to-date inventory of every sign — its restrictions, time conditions, applicable days, and precise location — has traditionally required teams of field workers manually recording each one. The process is costly, inconsistent, and outdated almost as soon as it finishes.

iRoadTech's parking module replaces that workflow with the inspection footage councils already capture. Drive the network once, and the AI handles detection, interpretation, geolocation, and publishing.

Live walkthrough of the Visualisation Portal — interactive map of detected parking signs.

How the module works

1. Video ingestion & frame extraction

Inspection videos are uploaded in standard formats (MP4, AVI, MOV, MKV) along with GPS trace data. The system extracts frames at configurable intervals and synchronises each frame with interpolated GPS coordinates — bridging the gap between 30 Hz camera capture and 1 Hz GPS logging.

2. Sign detection

Each frame passes through a vision language model (currently Qwen3-VL) that locates parking sign candidates. Detected regions are cropped and forwarded for detailed interpretation. The pipeline handles multiple signs per frame, varying lighting, and partial occlusions.

3. Sign interpretation

Cropped sign images are processed by a configurable multi-VLM pipeline. We support Claude (Anthropic), GPT-4o (OpenAI), Gemini (Google), and Qwen (Alibaba) — selectable per deployment. The VLM output is normalised and validated by a deterministic rules engine that enforces Australian parking sign standards (VicRoads TEM Vol 2 Part 211, AS1742.11).

{
  "time_limit": "2P",
  "time_range": "8:00am - 6:00pm",
  "applicable_days": ["Mon", "Tue", "Wed", "Thu", "Fri"],
  "arrow_direction": "left",
  "special_conditions": [],
  "confidence": 0.94
}

4. GPS mapping & road snapping

Each interpreted sign is placed on the map using its interpolated GPS coordinate, then snapped to the nearest road segment using the OSRM Nearest API with OpenStreetMap road network data. Markers sit on actual streets rather than at raw GPS positions that may drift into buildings or footpaths.

5. Visualisation & export

Results are published to the iRoadTech Visualisation Portal as an interactive Leaflet map with road-snapped markers, filterable by sign type, time-of-day, day-of-week, and street. Structured JSON exports plug straight into council asset management systems.

Demonstrated results

In our reference deployment with streets around Monash University Clayton campus — including Beddoe Avenue, Marshall Avenue, Stockdale Avenue, and Koonawarra Street — the system successfully detected and interpreted parking signs with high accuracy, producing a complete interactive map of parking restrictions.

Built for those who manage our roads

Pair it with