🛰️ Geocek — OSINT Visual Signal-to-Coordinate Engine

Geocek is a specialized tool for open-source intelligence (OSINT) research, designed to geolocate images based on visual cues such as license plates, road signage, infrastructure, and road layouts.

It employs a Constraint-Intersection Geolocation methodology, utilizing public OpenStreetMap (OSM) data via the Overpass API and Nominatim.

🌟 Features

• No Paid APIs: Uses 100% free OpenStreetMap-based services.
• Plate Prefix Mapping: Automatically narrows search areas based on regional license plate prefixes (currently focusing on Indonesia).
• Landmark Normalization: Handles fuzzy mentions of POIs (Point of Interest) like hospitals, malls, etc.
• Road Infrastructure Logic: Filters by road classification, lane count, and presence of medians.
• Confidence Scoring: Weighted scoring system to estimate location probability.
• Multi-Output Format: Generates text reports, GeoJSON data, and interactive Folium maps.

🛠️ Installation

1. Ensure you have Python 3.11+ installed.
2. Clone this repository or download the files.
3. Install dependencies:

   pip install -r requirements.txt

🚀 Usage

Generate geolocation candidates from an analysis JSON file:

python main.py -i tests/sample_input.json -f all -v

Options:

• -i, --input: Path to your visual metadata JSON file.
• -f, --format: Choose output between report, geojson, map, or all.
• -o, --output-dir: Where to save results (default: output/).
• -v, --verbose: Enable detailed processing logs.

🔍 Methodology

1. Signal Extraction: Visual cues are parsed into a SignalBundle.
2. Area Refinement: The search begins with a wide bounding box based on the license plate region, then refines via Nominatim geocoding of landmarks.
3. Overpass Query: Searches for matching nodes and ways (hospitals, roads) within the refined bounding box.
4. Constraint Filtering: Intersects candidates that meet multiple criteria (e.g., a hospital within 500m of a 4-lane primary road).
5. Scoring & Estimation: Assigns a confidence score and estimates a radius error based on signal matches and spatial proximity.

⚠️ Credits & Respect

This tool relies on the hard work of the OpenStreetMap community. Please respect the usage policies of the public APIs:

• Nominatim: Max 1 request per second.
• Overpass API: Use public instances sparingly.

---

Created as part of the Geocek internal toolsuite for OSINT research.