HighwayView

Overview

HighwayView is a toolkit designed to scrape and aggregate highway cameras from major European countries.

It handles scraping, aggregation, deduplication, verification, visualization, and DATEX II / CCISS traffic alert overlays for OBS Browser Source.

Supported Functionalities

• Scraping & Parsing: Extract camera data and media URLs from multiple European highway agencies into a unified structured JSON format.
• Verification & Deduplication (camera_check.py): Automatically check camera availability, filter out offline feeds, and remove visually duplicate images or error screens.
• Slideshow Generation (create_html.py): Create optimized HTML slideshows from the camera data. Features lazy loading, memory management, and error handling, making it ideal for streaming software like OBS.
• Curated Camera Loops (create_camera_loop.py): Automatically generate ~10-minute curated cycles of the most important national highways for each supported country.
• Data Inspection (list_cameras.py): Command-line utility to quickly summarize highway and camera counts from parsed datasets.
• DATEX II / CCISS Overlays (DatexParser module): Export traffic alerts for Spain, France, Italy, and the Netherlands to per-country OBS overlay data files. Netherlands DATEX II alerts are enriched from NDW VILD Alert-C location data.

Usage Examples

Ensure you have the uv package manager installed.

Run the Full Pipeline To run the main orchestration script that downloads, parses, checks cameras, and creates HTML slideshows:

uv run main.py

Generate HTML Slideshows for Specific Highways You can use the HTML generator to filter for specific routes and set a custom interval (e.g., Spain's AP-7 and A-7 with 10s intervals):

uv run tools/create_html.py data/cameras_es_online.json --highways AP-7,A-7 --interval 10

Generate HTML Slideshows for Specific Cameras Only include specific camera IDs from the UK dataset:

uv run tools/create_html.py data/cameras_uk_online.json -c cam_m25_1 cam_m25_2

Verify a Dataset Identify offline or broken cameras and output a clean JSON dataset:

uv run tools/camera_check.py data/france_original.json

List Camera Counts Print a formatted list of all highways and their valid cameras from a generated dataset:

uv run tools/list_cameras.py data/cameras_it_online.json

Serve OBS Browser Sources Locally Serve current generated camera and overlay artifacts through a local aiohttp server:

uv run tools/serve_obs.py

Use these URLs in OBS Browser Source:

http://127.0.0.1:8765/cameras/NL
http://127.0.0.1:8765/cameras/BE
http://127.0.0.1:8765/overlay/ES/

NL and BE camera media are proxied through the local server so HighwayView can attach provider Referer headers when fetching the actual camera feeds. BE slideshow iframes use /proxy/cameras/BE/?name=<camera_id>; the local server then proxies the BE player assets and HLS stream URLs.

DatexParser module examples

Generate Traffic Alert Overlays Generate overlay data for all configured DATEX II / CCISS countries once:

uv run get_datex.py --once

Generate a single country overlay:

uv run get_datex.py --country ES --once

Generate the Netherlands overlay from the NDW gzipped DATEX II feed:

uv run get_datex.py --country NL --once

Use a custom road whitelist:

uv run get_datex.py --country FR --roads A7,A16 --once

Disable heuristic filtering:

uv run get_datex.py --no-filter --once

Run continuously (refresh every 5 minutes by default):

uv run get_datex.py

Use the per-country overlay UI files:

data/overlay_spain/index.html
data/overlay_france/index.html
data/overlay_italy/index.html
data/overlay_netherlands/index.html

Netherlands VILD lookup

The Netherlands DATEX parser uses DatexParser/data/vild_6.13.A.sqlite to translate Alert-C/VILD LOC_NR values into road names, nearby areas, and administrative fields. The VILD value is exported as reference_marker; VILD hectometer ranges and DATEX offsets are used internally to estimate km_point. Raw alertc_* fields and VILD km_start_* / km_end_* ranges are not written to overlay_data.json.

Rebuild the SQLite lookup from the NDW VILD DBF when updating VILD versions:

uv run DatexParser/vild_dbf_to_sqlite.py path/to/VILD6.13.A.dbf DatexParser/data/vild_6.13.A.sqlite

Project Structure

The project is split into three modules and an orchestration script.

• main.py: Orchestrates the scraping, parsing, checking, and visualization of camera data.
• get_datex.py: Utilizes DatexParser to export DATEX II / CCISS traffic alerts to JSON files.
• config.py: Stores country-specific constants and shared settings.
• Downloaders/: Contains the scraping module for each country.
• Parsers/: Contains the parsing module for each country.
• DatexParser/: Contains DATEX II XML parsing, Italian CCISS parsing, Dutch VILD lookup generation, heuristic filtering, and overlay export code.
• tools/: Contains the tools for checking, visualizing, and locally serving OBS runtime artifacts.
• data/: Contains the raw and processed camera data plus per-country overlay assets.

Documentation

• Architecture: docs/Architecture.md — subsystem design, data flow, and component overview.
• API Reference: Generated from docstrings. Open docs/index.html in a browser.

Supported Sources

• France
  • Bison Futé
  • ASFA (Association des Sociétés Françaises d'Autoroutes)
  • DATEX II Feed
• Italy
  • Autostrade de Italia
  • Autostrada del Brennero
  • Autostrada Brescia Verona Vicenza Padova (ABP)
  • Concessioni Autostradali Venete (CAV)
  • SATAP (Società Autostrada Torino-Alessandria-Piacenza S.p.A.)
  • CCISS Feed
• Spain
  • DGT (Dirección General de Tráfico)
  • DATEX II Feed
• Netherlands
  • Rijkswaterstaat cameras
  • NDW DATEX II actueel beeld feed
  • NDW VILD location reference data
• Belgium
  • Vlaams Verkeerscentrum cameras
• UK
  • Highways England