Displaying the finished, positioned pictures is not the impressive part of photosynth, that is more or less standard polygon pushing (or could have been).
The unique point of photosynth is precisely the automatic spatial reconstruction based solely on image analysis. This is preprocessed, heavy-duty data cruching, and there is no point in trying to do it in Action scipt. Coding your own processor can certainly be done, since it already exists, but I think it is safe to say it would qualify as very hard work indeed. If you already _have_ a dataset prepared, then you could indeed display the images in a similar way with Away3D. But of course with the typical performance hit compared to native 3D, so you couldn't get the kind of volume of images that makes photsynth so immersive. With a carefully selected set of images and angles, I'm sure one could make an impressive smaller-scale presentation, though. On Mon, Mar 9, 2009 at 8:56 AM, [email protected] <[email protected]> wrote: > > http://photosynth.net/about.aspx > http://photosynth.net/explore.aspx > http://photosynth.net/view.aspx?cid=39e43cf8-8a7e-4a3c-8cd9-31ee91767f0f > > it is a cool effect. but if use the 3d engine. we also have two > question to do . > > first, we must can get the photo real-world-angle,size and zoom. > > second, we must set the relative-position at away3d system. > > is it hard to work? or nerver can be done ?
