PhotoSynth

That's phenomenal. Simply amazing.

I don't quite understand how it's doing what it is, but I think I can see the ramifications.

Awsome.
 
err, I'm still not really convinced of this Idea, I'd rather see someone, make 3d data from actual realtime satelite photos reconstructed in 3d with high res textures..... alas it is but a pipe dream.
 
Anhar Miah said:
err, I'm still not really convinced of this Idea, I'd rather see someone, make 3d data from actual realtime satelite photos reconstructed in 3d with high res textures..... alas it is but a pipe dream.
Click to expand...

I think this technology is a heading in the right direction though. If google earth got ahold of this similar thing I'm sure they could make your dream come true in the coming years. Although realtime satellite images won't be available to the public for privacy/security issues.
 
I'll believe it when I "really" see it. Presentation videos rarely show the real complexities. Only the polished finish. I mean, they didn't show any of the prep work, only the final results. And they were obviously using photos that were taken with the same camera, on the same day, by the same person, and most likely from the same spot (for better perspectives). And they are saying that it will place photos, scattered from around the Internet, and place it into its appropriate 3d space. Something tells me that task is much more difficult than it sounds.

PK (aka Party Pooper) :)
 
The idea that spiders would gather these images and actually look at them and analyze them and construct a common image out of them is a pretty bold statement, not sure if our computing power is up to it yet, nor our bandwidth to do it efficiently...
 
greeches said:
The idea that spiders would gather these images and actually look at them and analyze them and construct a common image out of them is a pretty bold statement, not sure if our computing power is up to it yet, nor our bandwidth to do it efficiently...
Click to expand...
I was thinking the same thing. First, I assume the presentation computer was very high end, and it was most definitely using Vista and its new graphics engine. But this is probably the most innovative product Microsoft has conceived in quite some time. It is certainly a creative idea--let's see how it "pans out."
 
The point is to link images together through common content, a huge idea, IMHO...

We think of links in text, Wikki, and what's its done in terms of knowlege, what about images in terms of seeing things that you cannot experience first hand....


It put a new wrinkle in my brain...
 
greeches said:
The point is to link images together through common content,
Click to expand...


Ands THATS the biggest technical problem I forsee, if you have a computer analyze pictures and try and connect them, you have a BIG BIG problems, if you have used OCR, or any form of "Recognition" then you will know how flawed it is.

I think there is a huge "nudge" factor going on, Like the 3d Geomotery is already precalculated and then the images placed in reference to them NOT the other way round (if they have done that then I'll eat my hat)


Anyway its amazing stuff, but I need more info on the process before I jump up and down...(I'm not diss'in it though)

Anhar
 
You must log in or register to reply here.
Back
Top