After the buzz has died down about the new sky feature in Earth, I personally think one of the most useful new features is the ability to place images in the landscape, matching the perspective of the location from which they where taken.
The new photooverlay element in KML 2.2 is used to great effect with the gigapxl images, which you just have to try out to really appreciate.
You can just keep on zooming into these amazing images..
And these are taken with a single exposure, so they are very useful for analysis of the images.
For me the most exciting development is the potential now for those millions of geotagged photos already online to be placed in the locations from where the pictures were taken, this would provide another great way of representing the “sense of place” of a location.
As a simple example, this is a picture of the Google Office in London, taken with the Camera in my Nokia N95. With its GPS, and the additional of a simple compass to provide the bearing of the shot, you have all your need.
Written and submitted from home, using my home 802.11 network.
2 replies on “Put your pictures on the planet..”
I can only assume, that not that far off into the future — and once a substantial amount of photographs can be catalogued —- that, Google Earth would be able to automatically texture buildings in this space, if at least the coordinates are accurate enough.
Or, at least if the coordinates are accurate to the point that they can auto-‘snap’ to the buildings.
I’m not sure it’d be entirely possible — but the concept itself seems feasible enough to assume. Then again, I suppose the Streetview cameras would be better suited for that approach.
PS:
If you were able to ‘auto-snap’ in this manner by angle and points of the buildings — to auto-texture… Something tells me that you can re-channel your approach to Streetview for Maps, and the actual rich immersion can occur within Google Earth as a more ‘true’ visualization experience.
I can think of endless reasons why this would be the ultimate approach to nearly every end-user application.
The government doesn’t even have that kind of visualization under its belt — and I’m certain they’d love to have it. Mix that in with robotics and UAV technologies, and what you have is the ultimate system for visualization prior to actually setting foot in a location.
Of course, you could consider the same approach with video registration, but I don’t like that approach since you’re talking low-res. Though, the low-bandwidth model of video would be incredibly interesting for various end-uses as well.