An AI-powered web application for digitizing and analyzing zoning maps from PDF documents. This tool streamlines the conversion of static zoning PDFs into interactive, editable vector data for urban planning and GIS workflows. Bring your own API key and you're off to the races.
- GPT-4 Vision Integration: Automated map feature detection and overlay alignment
- Smart Legend Detection: Automatically extracts zoning types, codes, and color mappings
- Intelligent Coordinate Suggestion: Georeferencing using detected streets and landmarks
- Auto Alignment: AI-optimized scale, rotation, and positioning
- Drag-and-drop upload (max 10MB) with validation
- High-quality conversion with
PDF.js - Upload progress tracking and visual feedback
- Smart cropping to isolate map and legend content
- Seamless
OpenLayersintegration - Real-time controls: scale, opacity, rotation, position
- Interactive map with
OpenStreetMapbase layer - Live mouse position and scale indicator
- Drawing modes: Polygon, Rectangle, Circle, Freehand
- Assign zone types with color-coded visualization
- Edit properties of drawn features
- Includes 17+ predefined zoning types
- GeoJSON for web GIS compatibility
- KML for Google Earth or legacy mapping tools
- Summary reports with stats and metadata
- Styled exports maintaining zoning visual cues
- Upload: Drop your PDF zoning map
- Smart Crop: AI separates map and legend
- Georeference: Position map with coordinate suggestions
- Legend Detection: Zoning categories identified
- Zone Configuration: Manage zone types, colors, and rules
- Draw & Edit: Use advanced drawing tools for boundaries
- Export: Download in GeoJSON, KML, and more
- React 18 + TypeScript
- Vite (build + dev server)
- React Router DOM (routing)
shadcn/ui+ Radix Primitives- Tailwind CSS
- Lucide Icons
- Sonner (toasts/alerts)
OpenLayersfor high-performance mappingPDF.jsfor client-side PDF parsing- GeoJSON for standard vector format
- GPT-4 Vision (via OpenAI API)
- Custom vision models for:
- Feature detection
- Geocoordinate inference
- Legend parsing
- Map alignment
@tanstack/react-queryfor data fetchingReact Hook Form+ Zod for forms- Supabase (PostgreSQL) for persistence
- Custom hooks for map feature state
- TypeScript
- ESLint
- Tailwind typography plugin
- PostCSS + Autoprefixer
- Urban Planners: Convert legacy PDFs to digital zoning data
- GIS Professionals: Generate usable shapefiles and layers
- Municipal Governments: Modernize zoning documents
- Real Estate: Visualize property zoning at parcel level
- Consultants: Automate zoning workflows with AI
# Clone the repository
git clone <repository-url>
cd zoning-map-digitizer
# Install dependencies
npm install
# Start dev server
npm run dev
# Build for production
npm run build- Requires OpenAI API Key
- Requires Supabase Project (for data backend)
- Use
.envfile to configure API keys and Supabase connection
This application bridges traditional GIS workflows with modern AI capabilities, reducing the time, effort, and manual steps involved in zoning map digitization.
feat: Allow toggling of PDF overlay #1 by @weissbc07
- @weissbc07 made their first contribution in #1
Full Changelog: Commits »
