Thank you Joojaa.
What I really want to achieve is the set lighting reconstruction (highest priority) and maybe 3D set reconstruction (for reflection, lower priority and not necessary). Since I will shoot a HDR panorama for IBL anyway, that's why I was asking if there is a way to derive locators with spatially accurate measurement for the scene features in either Photogrammetry or tracking applications.
So, would you mind providing me some suggestions for simplest but accurate approach for set lighting reconstruction? I read SynthEyes' manual and realized that it has a "multi-shot tracking" feature where you can combine multiple footages to build coordinate system and hence reconstruct the set accurately. I am wondering if I can leverage "ready-made" panorama (for IBL as mentioned before) plus some still images or video footages shot on set (of course with parallax this time), and couple tape measurements taken on set to achieve the set lighting (or even 3D) reconstruction? Or is there a simpler approach with fewer images or video footage to take?
QUOTE(Joojaa @ 04/19/09, 10:13 AM) [snapback]306186[/snapback]
The single spherical panorama has no parallax detail thus theres nothing to measure.
However if you have 3 spherical panoramas offeet form each other then something may be measured*. It is foccourse better the more offset frames you have as you can even out measure errors.
But one picture no parallax no information no depth, nothing to track and triangulate. This is by the way why humans have 2 eyes.
*2 is technically sufficient but because the panorama does become in line in the offset direction 3 points not in line would be better or youd loose one line.