Hi, well, maybe someone else can help with the automated CPs finding, because I really dont use it. I usually put them manually as I also has never liked the automated results.
What I can see is that I have reduced the biggest CP distance to 14 doing a "everything without translation" optimization. After looking to the CPs table (the buttom beside the preview one) and listing CPs in order of distance I looked at CP #1, which has the biggest distance. The image really looks like having some little parallax problem, but I don't really know how does it affects the auto CP finding. I usually also don't like CPs over trees and sky, as they can move from one shoot to the other. I guess there is a way to filter CPs in the sky (celeste) but don't know about automatically remove them from trees. I with we could have an option to specify regions where we would like CPs. Maybe with this an automation would be easier. I am almost sure many people do panoramas automatically. In my short experience, even when I generated CPs automatically and used them, I had to manually remove some, so the automation would not be possible. That's why I still do CPs manually. After writing up to this point I decided to try another thing: remove the CPs you added manually (20, 21, 22) and reset the project to execute the optimization I usually do. I got better results, so, if you have obtained the points 1-19 automatically then you can automate it :) I usually do optimizations progressively. I first do a Position (y, p, r), and after this I do a (y, p, r, v) and to finish I do an "everything without translation". Doing this in your project gave me a max distance of 5 pixels and a good final image. I put the project attached. Cheers, Carlos E G Carvalho (Cartola) http://cartola.org/360 2012/1/3 Gerhard Killesreiter <gerh...@killesreiter.de> > -----BEGIN PGP SIGNED MESSAGE----- > Hash: SHA1 > > Am 03.01.2012 12:27, schrieb kfj: > > On 3 Jan., 12:08, Gerhard Killesreiter <gerh...@killesreiter.de> > > wrote: > > > >> I've tried adding CPs manually but the matching process wasn't really > >> working. > > > > Do you mean that you had difficulties placing the control points, or > > that the optimizer failed to produce a useful result? > > The optimizer did add usefull CPs after some changing of parameters, but > these are clustered in the mid-section of the images. In order to get a > working pano I needed to add additional CPs in the foreground manually. > > The optimiser was usually not able to find the point I had added in the > other image and I had to add it manually there too. > > After omtimizing the optimizer mostly complained that the distance was > too big. > > I even got hugin to crash once... > > I was able to stitch in the end, but I don't think it worked out very well. > > I suspect that I may have a bit too much parallax error... > > It is possible that there's an issue with the exif data too, since my > camera doesn't write the correct data. > > I then enter the data manually: > Focal lenght: 6.75 > crop factor: 1.5 > > (that mask could use some explanation which focal length is needed: the > real one or the 35mm equivalent. I think it's the latter). > > >> My concern is that I either used the wrong technique to take the > >> photos, my workflow is somehow wrong, or the lens has a defect. > > > > I suppose there's little you can do wrong taking the images, and I > > also think a lens defect isn't very likely. Leaves your workflow ;-) > > Yeah, my suspicion too. > > >> Also, I really want to automate this as much as possible as I plan to > >> take lots of panos. > > > > If you can put your images and PTO online somewhere, I'll have a look > > if I can see what's going wrong. > > That'd be great! I'll send you a link in private. > > Note that the images are just test images, the idea is not to make a > nice pano from them, the idea is to understand why I don't get CPs in > the front automatically and what I can do to change that. > > In this case I have added the CPs 20-22 manually. > > > Cheers, > Gerhard > -----BEGIN PGP SIGNATURE----- > Version: GnuPG v1.4.10 (GNU/Linux) > > iEYEARECAAYFAk8DAxMACgkQfg6TFvELooQB9wCfQlWCIg9KnLw+LoJG0zUHfOqo > 0uUAnRAKn8x1Mp8A/gU3QxR12PScLa4J > =c0bg > -----END PGP SIGNATURE----- > > -- > 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 > -- 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
IMGP4921-IMGP4922-cartola.pto
Description: Binary data
