On 5 Mrz., 01:17, xraycrystallographer <[email protected]> wrote:
> This helps! I think this means that I need to do an accurate lens > calibration--I assume > the instructions in the tutorial are still valid? I am in the process of > building a > panorama head (based on a theodolite base) and recognize how important > camera position is. Let me propose a quick way for calibrating: make a full spherical 360X180 panorama with no parallactic errors (you'll have to have your NPP centerd) and 50% overlap. Do it indoors in a setting with structured content all around you (try ornate buildings). Then find CPs with autopano-sift-c with --maxdim 6000 --maxmatches 0 --projection %f,%v %o %i (assuming 6000 is bigger than your biggest image dimension) or, alternatively, use cpfind with --fullscale you'll get absolute tons of CPs, like literaly thousands with apsc and hundreds with cpfind. Optimize everything but translation, clean your CPs and optimize again. The a,b,c,d,e and v values will be very close to optimal and you can save that as a lens profile. Note that if your lens isn't fixed to the body, with every take you'll have to reoptimize d and e, but a,b and c will not change overly much. You may choose not to optimze c and fix it at zero. Keep in mind that if you're using a fisheye there is no NPP but only a point of least parallax for a given angle between images - meaning you can't ever get it 100% right. But if you're 99.9% there, the blender will sort out the rest. > This brings up another question. With my home-built > panorama head I can read the yaw to better than 30" of a degree. I note > that in general the scatter in the yaw angles differs by quite a bit more > than I would expect--some times +/- 1degree or more from the actual angle. > Would I be better off fixing the yaw at the set angle? I realize there is > probably some slop in things like the lens fitting on the camera etc., etc., > etc., but I should be able to fix most of these. I'm also using a homemade pano head, and I don't ever make an effort to be very precise about the angles - the optimizer sorts it out as long as you have enough overlap and your calibration is good. The scatter should lessen with good calibration - once you have that it's - as I said - just nudging the correctly transformed data into place. Kay -- 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 [email protected] To unsubscribe from this group, send email to [email protected] For more options, visit this group at http://groups.google.com/group/hugin-ptx
