Ah yes. Maybe I gave too much info on the particular image and confused my question. It's probably much simpler than folks were probably thinking.
I shouldn't have said "fixing the horizon manually," as I'm not really concerned about trying to make the picture look like it was taken in one go. I just want to make decisions on what is in the picture and how it is blended. I actually like the fact that the horizon is moving as it shows in a single image the dynamic nature of the situation in time and space. So to pose my problem more clearly, I should have simply said that I want to "adjust the image manually." I am interested in what methods people are using for making easy and efficient "hand adjustments." My particular interest is about how people choose to output the images for ease of registration and masking, and also what approach they take to doing that masking and adjustment. Do people use the individual remapped images and manually place them in a panorama sized image document? Do they do what I posed and output just two panos of alternating remapped images? (I thought this might be easier since all photographic information could be contained in just 2 equal sized panoramas.) Are alpha channels the way to go making the blends and masks? Is there some other better way? tricks tips etc. Thanks --John On Nov 4, 3:40 am, kfj <[email protected]> wrote: > On 4 Nov., 06:07, JohnPW <[email protected]> wrote: > > > I have some images I recently took handheld on a small yacht sailing > > in San Francisco Bay. I had to put all the control points on > > stationary parts of the boat since anything else was in constant > > motion. It has made me want to do more hand blending and masking to > > make the (very mismatched) horizon less jarring and to fix a few > > things on the boat that moved. I was considering how best to do it. > > Forgive my slightly off-topic reply, but here it is anyway: > > When I was a student, I attempted a project to analyze some (real- > world, messy, processed...) animal feed raw materials by way of an > image analysis system. This was in the eighties: I had a (nice :-) > video microscope and a 512X512 monochrome frame grabber in an AT > compatible computer. I showed the images I was working on to the prof > who was reading computer vision at my uni. He looked at them and told > me my problem was 'ill-posed'. I was a bit miffed at the time, but as > my project proceeded I had no choice but eventually to come to the > same conclusion. > > If the scene isn't static, you'll have a very hard time 'stitching' it > together: it's like trying to make a garment from wrongly-cut pieces > of material. I'd say the problem is ill-posed. What you need isn't a > panorama stitcher but an optical system that captures all the moving > bits in one go. > > Yet I'm curious to hear a few good tips on fixing the horizon > 'manually', even though I usually manage to get the horizon right, > sacrificing precision elsewhere in the stitch. > > 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
