I hope you are trying to put the camera in the same relative position to do each shot, but maybe if you think it is not too precise it could be an idea to set different lenses for each image, so hugin can optimize them separately. I would do that after calibrating the lens, as suggested. I would also try to do the optimization little by little, starting with few items, maybe y, p, r only and then going through other parameters.
If you get desperate you can try to make your images available (maybe resized) so we can try to make it. Good luck, Carlos E G Carvalho (Cartola) http://cartola.org/360 2012/1/30 kfj <[email protected]> > On 30 Jan., 08:45, torger <[email protected]> wrote: > > Yes I have looked at that. They use a dummy fov and just skip yaw/ > > pitch optimization. I don't get any good fit then. I think the problem/ > > difference here is that I'm not using a flatbed scanner but have a > > real lens, that is there is not zero distortion. It may actually also > > be some very small yaw and pitch since it might not be 100% exactly > > perpendicular. So probably there should actually be some small yaw and > > pitch to get the best fit. > > Reading your posting, I have the impression your camera is mounted to > a repro stand (I hope that's the right english word...). > > First calibrate your lens and setup. Find a subject which is rich in > non-repetitive detail (best would be photos of detail- and contrast- > rich natural subjects). Use your repro process on it - a few > additional images won't do any harm. Use the correct fov. Input > projection rectilinear, output projection rectilinear. Generate lots > of CPs, try cpfind with > > --fullscale --sieve2size 5 -o %o %s > > in the settings. Now optimize for x,y in all but the anchor and a,b,c > in all images. Note that you don't optimize for z, since with your > setup z is fixed. This should already deliver a reasonable result, but > in case you want to adjust to minimal errors in your setup you can > proceed to try this: > > Optimize for y, p and r only if you're not happy with the result as it > is. If you used a repro stand, all y, p and r values should be quite > close, and you can just average over the lot and use the average > values for all images in the future. > > If you didn't use a repro stand for the images, you will definitely > need to optimize for z, y, p and r and you're entering tricky > territory, since the optimizer now has to deal with (often too) many > degrees of freedom. Your mileage will vary, and the procedure for this > situation usually involves optimizing only a few variables at a time > and doing several iterations. Best to be avoided, laborious and not > repeatable, and you may have to add line control points to get > anywhere. > > Once you've gone through the calibration process, you have > > - lens calibration data which you can save in a lens.ini file to use > for future projects with this lens > - possibly y, p and r values for your mount which you can also reuse > if the mount is unchanged > > You may want to correct for lens shift as well, but it's tricky doing > that with x and y being optimized at the same time - nevertheless you > may want to experiment with the lens shift parameters. You might try > and switch off x and y in the optimization when you optimize for d and > e, and vice versa. If you alternate between the two for a few > iterations, you may arrive at a reasonable result. > > For future stitches with your repro setup, you can now simply load the > lens correction data from your saved lens.ini file, set y, p and r, > and only optimze for x and y only > > So, once again in a nutshell > > - use a repro stand if you can > - calibrate your lens > - optimze y, p and r only if you must. > > 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 > -- 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
