## Re: [PanoTools-devel] Improving PTblender?

 Re: [PanoTools-devel] Improving PTblender? From: Pablo dAngelo - 2007-03-16 11:57:13 ```"Erik Krause" wrote > On Friday, March 16, 2007 at 8:24, Pablo d'Angelo wrote: >=20 > > Original scene: > > http://grail.cs.washington.edu/projects/vignette/origs/greenlake=5Fap=5Fso= lved.jpg >=20 > should read=20 > http://grail.cs.washington.edu/projects/vignette/origs/greenlake=5Fap=5Funso= lved.jpg > of course. True. A mixup on my side. > > Correction by Goldman and Chen (best algorithm in the literature I kno= w) > > http://grail.cs.washington.edu/projects/vignette/origs/greenlake=5Fap=5Fso= lved.jpg > >=20 > > Correction with my new algorithm > > http://home.arcor.de/pablo.dangelo/tmp/green-lake=5Fvig=5Fresp=5Fexp.jpg > >=20 > > All corrected fully automatically, no manual tweaking of parameters in= volved. >=20 > Looks very promising, although there is a color shift in the leftmost=20 > part. Later I can also post an image where no color/white balance correction has been done. For the current example, the white balance on the rightmost= image was used as a base. Additionally with the algorithm it is also possible to use multiple "white balance anchor" images, to avoid the accumulation of errors over a large chain of images. > It would be interesting to see a comparison with an enblend=20 > result (with and without prior correction by your algorithm). I'm currently interested in improving the correction algorithm, so looking= at the enblended result would be of little value. For a final panorama, obviously enblend should be used. ciao Pablo =5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F= Erweitern Sie FreeMail zu einem noch leistungsst=E4rkeren E-Mail-Postfach! =09 Mehr Infos unter http://freemail.web.de/home/landingpad/=3Fmc=3D021131 ```

 Re: [PanoTools-devel] Improving PTblender? From: Pablo dAngelo - 2007-03-16 11:57:13 ```"Erik Krause" wrote > On Friday, March 16, 2007 at 8:24, Pablo d'Angelo wrote: >=20 > > Original scene: > > http://grail.cs.washington.edu/projects/vignette/origs/greenlake=5Fap=5Fso= lved.jpg >=20 > should read=20 > http://grail.cs.washington.edu/projects/vignette/origs/greenlake=5Fap=5Funso= lved.jpg > of course. True. A mixup on my side. > > Correction by Goldman and Chen (best algorithm in the literature I kno= w) > > http://grail.cs.washington.edu/projects/vignette/origs/greenlake=5Fap=5Fso= lved.jpg > >=20 > > Correction with my new algorithm > > http://home.arcor.de/pablo.dangelo/tmp/green-lake=5Fvig=5Fresp=5Fexp.jpg > >=20 > > All corrected fully automatically, no manual tweaking of parameters in= volved. >=20 > Looks very promising, although there is a color shift in the leftmost=20 > part. Later I can also post an image where no color/white balance correction has been done. For the current example, the white balance on the rightmost= image was used as a base. Additionally with the algorithm it is also possible to use multiple "white balance anchor" images, to avoid the accumulation of errors over a large chain of images. > It would be interesting to see a comparison with an enblend=20 > result (with and without prior correction by your algorithm). I'm currently interested in improving the correction algorithm, so looking= at the enblended result would be of little value. For a final panorama, obviously enblend should be used. ciao Pablo =5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F=5F= Erweitern Sie FreeMail zu einem noch leistungsst=E4rkeren E-Mail-Postfach! =09 Mehr Infos unter http://freemail.web.de/home/landingpad/=3Fmc=3D021131 ```