2009/9/19 Harry van der Wolf <hvdw...@gmail.com>: > > > 2009/9/19 Lukáš Jirkovský <l.jirkov...@gmail.com> >> >> Hello everyone! >> >> From today I'm finally quite content about state of the deghosting >> project. Everything is tested and I don't see any major problems with >> it. With this mail I ask everyone willing to test some "new" code. >> >> What I need most: >> >> Compilation on frequently used platforms: >> Linux – it works here on Archlinux so I guess that it should work in >> other distributions but I'll still welcome reports from other >> distributions >> Windows – with Thomas' help it have been made to compile on widows >> Mac OS X – I don't know about any attempts to compile it on OS X > > It compiles fine on OSX using cmake, already for quite some time. I never > mentioned it or tested it as you were continuously developing (I don't like > chasing a moving target) and, off course, I had this 6 weeks "black hole" > when I didn't have a mac. > > I just added it to the Xcode project. It took some "Sherlock Holmes" like > detective work before I understood that the complete old deghosting folder > had been "replaced". I can't add this XCode changes to the > gsoc2009_deghosting trunk for merge later on as that doesn't work. I have to > redo this when when your work is merged to the development trunk (or 2009.4) > but that's no problem. Now that I've worked it out and written it down it's > 10 minutes work.
Thank you very much! > >> >> Testing hugin_hdrmerge: >> Windows and OS X testing is needed. Since I've completely rewrote >> deghosting implementation it would be nice to test it. But please, do >> not consider different results as regression. There are also some >> other fixes (eg. avg_slow should work now) > > Since you've been working so hard on this, I assume you have one or more > test sets with final images/panos (with/without deghosting). Could you make > them available to us? In this case we can compare results and track down > where it functions well and where not (and why). I don't have enough time today (going out in a while) but I'll make (and upload) some nice set tomorrow. > >> >> Testing deghosting_mask: >> I hope the usage text is self explaining. Area where ghosting appear >> should be black in resulting masks. These masks can be used with >> enfuse-mask to allow deghosting in enfuse. Unfortunately it doesn't >> work with small artifacts (like moving branches). I hope I'll be able >> to improve this in future. To be honest integration directly into the >> enfuse has much worse results than this but I've different approach in >> mind. >> > > Again: Do you have test sets? > > >> >> final note: >> Yes, I know that it's slow. You can get noticeable speedup by passing >> -a f option to both deghosting_mask and hugin_hdrmerge. This will turn >> off using processing images as color but it also results in worse >> results. This option is similar to the old deghosting code. >> >> thanks for your time, >> Lukáš >> > > Hoi, > Harry > > > > Lukáš --~--~---------~--~----~------------~-------~--~----~ 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 -~----------~----~----~----~------~----~------~--~---
