Forum Discussion
The Foundry just solved live action stereoscopic VR 360
http://www.cgchannel.com/2015/04/the-fo ... f-vr-work/ :!:
There are some innovations there but not fundamental solutions. Tracking --- it can solve camera positions, orientations and internal parameters of a ring of Gopro cameras. This is good but not unique. Photoscan3d (pro version), for example, if their product description is accurate, can solve fisheye array images like this (and do a scene reconstruction and create a panorama from a nodal subset of images).
http://www.agisoft.com/features/compare/
Also in the Nuke process it seems to solve the cameras from scratch each time whereas a more trustworthy process and one that would be better for creating reusable templates maybe would be to determine the lens/sensor parameters of each camera separately (Gopros vary a lot, for example in the positioning of the sensor behind the lens). You can do this calibration with an accurate indexed panorama head and PTGui or Hugin. It would be nice to be able to input this information into Nuke (or use Nuke to do this individual calibration via panorama head) -- and then go from there to calibrate the positions and orientations and fine tune the individual camera internal parameters of the camera array.
Then he talks about stitching. He shows how each of the stereo panoramas from the L/R pairs have geometric -- ie parallax (and sync presumably with this rig) stitching errors which can be corrected automatically by Nuke's warping capabilities. But the L and R stitching errors (speaking parallax errors only) will likely be in different locations in the L and R panoramas and there is no guarantee (at least he didnt mention that they are handling this) that the blending algorithms will act in stereoscopically identical fashion.
Then he talks about how correct stereo for zenith and nadir is impossible without a textured mesh of scene geometry (or by using depth maps to reduce nadir and zenith depth to zero -- infinity in vr headset contexts). In fact Micoy at least claims correct nadir and zenith stereo viewing is possible (they sell plugins for full dome stereo rendering and have patents for a spherical camera array where images are stitched in spiral patterns).
http://www.micoy.com
http://www.mtbs3d.com/index.php/article ... aging-Tech
http://www.trentgrover.com/images/tech/ ... oyComp.pdf
(the Trent Grover article has swapped captions for his zenith anaglyph stitching illustrations I think)
Also the Iximage sector approach to "omnipolar" fisheye image stitching seems to me to have distinctly superior nadir/zenith stereo compared with normal camera array geometry rigs (rings or spheres of cameras).
http://www.agisoft.com/features/compare/
Also in the Nuke process it seems to solve the cameras from scratch each time whereas a more trustworthy process and one that would be better for creating reusable templates maybe would be to determine the lens/sensor parameters of each camera separately (Gopros vary a lot, for example in the positioning of the sensor behind the lens). You can do this calibration with an accurate indexed panorama head and PTGui or Hugin. It would be nice to be able to input this information into Nuke (or use Nuke to do this individual calibration via panorama head) -- and then go from there to calibrate the positions and orientations and fine tune the individual camera internal parameters of the camera array.
Then he talks about stitching. He shows how each of the stereo panoramas from the L/R pairs have geometric -- ie parallax (and sync presumably with this rig) stitching errors which can be corrected automatically by Nuke's warping capabilities. But the L and R stitching errors (speaking parallax errors only) will likely be in different locations in the L and R panoramas and there is no guarantee (at least he didnt mention that they are handling this) that the blending algorithms will act in stereoscopically identical fashion.
Then he talks about how correct stereo for zenith and nadir is impossible without a textured mesh of scene geometry (or by using depth maps to reduce nadir and zenith depth to zero -- infinity in vr headset contexts). In fact Micoy at least claims correct nadir and zenith stereo viewing is possible (they sell plugins for full dome stereo rendering and have patents for a spherical camera array where images are stitched in spiral patterns).
http://www.micoy.com
http://www.mtbs3d.com/index.php/article ... aging-Tech
http://www.trentgrover.com/images/tech/ ... oyComp.pdf
(the Trent Grover article has swapped captions for his zenith anaglyph stitching illustrations I think)
Also the Iximage sector approach to "omnipolar" fisheye image stitching seems to me to have distinctly superior nadir/zenith stereo compared with normal camera array geometry rigs (rings or spheres of cameras).