Hi Dan & Max I just finished a careful analysis of the differences between the equiToCentralToPanini you posted earlier, and my projection of it with Panini; and now you had to go and get it right ;-(. I won't bother to post my images because they are identical to your new one, up to a linear scale factor on the Y axis.
Max's are the same, too, up to scaling on X and Y (which is also possible in Panini). So it seems that Recti-Perspective Projection == Panini Projection, and we all know how to do it right. This is a Good Thing. Regards, Tom PS how did you make that 360 with the straight back walls?? With Panini I can only go up to 309 degrees wide because my viewing projection fov is limited to 150 degrees, which sees that much of the inside of the circle at an eye distance of 1.07 radii. But past 220 degrees hfov everything looks curvy, not straight. On Jan 2, 9:43 pm, "Daniel German" <d...@uvic.ca> wrote: > On Fri, Jan 2, 2009 at 6:25 PM, Max Lyons <max.ly...@verizon.net> wrote: > > > Dan, > > > Great picture! A perfect location for this type of projection. > > Thanks Max. > > I think I finally go the original panini right: > > http://turingmachine.org/~dmg/temp/equiToCentralCylindricalToPaniniCo... > > it looks very close to your 1:2. > > I am using mathlab so I don't have very good control of the field of > view. But I am going to try > see if I can fix libpano. > > --dmg > > > > > As Tom pointed out earlier today, I have added a number of new projections > > into > > PTAssembler, one of which ("recti-perspective") appears to be quite similar > > to > > the Panini projection. That said, here's a page with a few examples I > > generated in recti-perspective projection using your image (about 235x120 > > degrees). > > >http://www.tawbaware.com/forum_posts/recti_perspective.htm > > > I think that we must have a slightly different projection formulas because > > our > > results look a little different, although they seem quite similar in the > > central regions. > > > Max > > >> Hi everybody, let us get our understanding of what the projection is > >> supposed to do first. > > >> I am confounded by a lack of a good test case. So here is an image > >> that I suggest we should use to test our "versions" of Panini. > > >>http://turingmachine.org/~dmg/temp/equirectangular.jpg > > >> It is 3000x1500 pixels. I understand the geometry of this building > >> well, and there are only straight lines in it, with only 90 degrees > >> angles. the sides of the staircase run parallel to the side walls. The > >> far away walls and the back walls are all orthogonal. The area where > >> the tripod was placed is below a lower ceiling (appears very curve in > >> the image). > > >> Here is my version of Panini: it is 226 x 123 degrees field of view. I > >> am projection from the equirectangular, to cylindrical, and then to > >> stereographic. > > >>http://turingmachine.org/~dmg/temp/equiToCentralCylindricalToPanini.jpg > > >> I really like it. The only area where it breaks down is on the roof of > >> the ceiling above the tripod. > > >> Can all of you generate your versions, also at a width of 3000 pixels > >> from this image? > > >> --daniel > > -- > --dmg > > --- > Daniel M. Germanhttp://turingmachine.org --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "hugin and other free panoramic software" group. A list of frequently asked questions is available at: http://wiki.panotools.org/Hugin_FAQ To post to this group, send email to hugin-ptx@googlegroups.com To unsubscribe from this group, send email to hugin-ptx-unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/hugin-ptx -~----------~----~----~----~------~----~------~--~---
