RE: Video Geotagging Protocol For Electronic Maps: Concept

Hi Rob,

Your proposal looks quite interesting. Also, I like your application.  In the Maps for HTML Community Group, we encourage developers to bring ideas like this forward, in an attempt to identify commonalities which could form the basis of standard(s) (media types) which could / might be supported natively by Web browsers.  To do this, we’ve developed “Map Markup Language”, or MapML, which is a text-based hypermedia format for maps.  Although its early days for features-based MapML [1],[2], I can see the parallel between features and WebVTT  for making maps accessible.  Maps are similar to video in some ways, in that the view changes.

Anyway, not to stray too far from your proposal, I can see almost any HTML element having a geospatial location, although your proposal seems to target embedding location information on a per-frame basis.  How do you see such a video being used ?  What would the user experience / interface for those georeferenced frames be?  I am interested in how this might look encoded or used in MapML ☺.

[1] http://geogratis.gc.ca/mapml/client/map-carte.html

[2] https://maps4html.github.io/MapML/spec/


Also, FYI we are sponsoring OGC Testbed 14 especially for development of feature encodings for MapML: http://www.opengeospatial.org/standards/requests/160


Thanks,
Peter


Peter Rushforth

Technology Advisor
Canada Centre for Mapping and Earth Observation / Earth Sciences Sector
Natural Resources Canada / Government of Canada
peter.rushforth@canada.ca<mailto:peter.rushforth@canada.ca> / Tel: 613-759-7915

Conseiller technique
Centre canadien de cartographie et d’observation de la Terre / Secteur des sciences de la Terre
Ressources naturelles Canada / Gouvernement du Canada
peter.rushforth@canada.ca<mailto:peter.rushforth@canada.ca> / Tél: 613-759-7915

From: Rob Smith [mailto:rob.smith@awayteam.co.uk]
Sent: January 5, 2018 10:34 AM
To: public-sdwig@w3.org
Subject: Video Geotagging Protocol For Electronic Maps: Concept

Introduction
I've just had a telecon with François Daoust, who suggested posting this message to the Spatial Data on the Web Interest Group.

Concept
This proposal is for a simple text-based protocol to geotag video, so film footage can be synchronised with electronic maps, e.g. Google Maps. However, it offers more than just basic geolocation by also including presentational features.

Why Is It Needed?
The original inspiration was an enquiry from Ordnance Survey, who wanted help to present their map data alongside footage for a TV programme. It wasn’t possible within the extremely short timescales required, but I realised that there was a need and it prompted me to research the idea further. There are a number of ways to include geolocation data within MPEG (Moving Picture Experts Group) movies, but different manufacturers, e.g. Sony and Panasonic, have taken different approaches so there is no standard: some use metadata and others encode it at frame level. Even if such a standard existed, it would neither be particularly accessible in the encoded stream nor include presentational and annotative elements like map zoom level and labels for points of interest, which would be useful to animate the display and direct the viewers' attention to particular map features.

Away Team’s expertise in development using web-based map APIs (Application Programmer Interface), combined with our knowledge of subtitling YouTube videos, suggested an open format to add standardised map data to videos in a very accessible way. The idea uses a time format similar to W3C’s WebVTT (Web Video Text Track) to implement commands which drive common features of electronic maps such as drawing a line, placing a marker, adding a label, and framing the map data, without being specific to a particular API. As with WebVTT, the format is text based and simply synchronises the additional data, i.e. map or subtitle, with the video playhead time.

I developed an online service called Trkd™ Maps, which overlays standard GPX (GPS Exchange) files on maps in a web browser, without uploading. It supports the three most popular map APIs: OpenSpace, Leaflet and Google, so I’m pretty familiar with their common features and differences. It even includes a hot-swap feature to change API on the fly, so you can load your GPX file with OpenStreetMap data (in Leaflet) and then switch to Google Maps to use Street View without reloading, which demonstrates their commonality.

http://www.awayteam.co.uk/index.php/services/maps/about


Who Would Use It?
Mobile video devices are a growth market, with the increasing use of dash cams, drones, and body and helmet cams. Being able to geotag their video data, either in real-time or retrospectively, would add valuable additional context to the footage, so images seen in the movie could be quickly linked to features shown on the associated map presentation.

The uses of this are many and varied, from security and rescue services to entertainment. Coastguard and mountain rescue services use drones to search inaccessible areas for people in distress, so geolocation information would immediately show rescuers their location and the surrounding terrain. The police and coastguard use helicopter-mounted cameras for surveillance and reconnaissance, with pictures relayed to remote operations centres where synchronised geolocation data could give a vital edge to the decision makers.

Drones are also used to cover sports events in isolated or inaccessible areas, such as rallying, sailing and cycling, where competitors relative positions are important and geolocation data would allow remote monitoring for better accuracy. For those who prefer to participate, helmet cams can capture their adventures on snow, bike or parachute, where geolocation data would pinpoint their precise route, speed and distance travelled from the changing location over time.

While the professionals rely on expensive equipment, a modern smartphone typically includes a video-capable camera and a GNSS (Global Navigation Satellite System) receiver, so anyone with a mobile can capture all the data required cheaply with kit they already have. Creating a simple, accessible standard to synchronise video and geolocation provides the critical link to enable them to relive their exploits in a web browser or post it online for others.

Geotagging still photos is already commonplace, using the Exif (Exchangeable Image File Format) standard. This proposed protocol could do the same for video, and in a way that complements and enhances existing standards rather than competing against them.

Conclusion
I hope this provides a good outline of the idea. I'd like to gauge interest from the Spatial Data on the Web Interest Group, and would welcome constructive feedback regarding possible uses and feature requirements. The concept has already evolved considerably since it was conceived last year, and I'm keen to maintain that momentum and improve upon it with the benefit of the Group's experience and expertise to help develop a useful and versatile tool for the wider online community.

Rob Smith
Away Team Software Ltd
http://awayteam.co.uk/