Loading ...
Sorry, an error occurred while loading the content.

Re: [PanoTools-devel] [PanoToolsNG] Re: pano12 future development

Expand Messages
  • Daniel M. German
    Jim Watters twisted the bytes to say: ... Jim The thing I remember most of PTStitchers color and exposure correction Jim was how much it improved for my
    Message 1 of 1 , Jul 29, 2006
    • 0 Attachment
      Jim Watters twisted the bytes to say:

      Jim> dmgalpha wrote:
      >> Use PTblender and output the colour correction curves, then load them
      >> into photoshop or the gimp. This will give you the flexibility you
      >> need. You can run PTblender with only the images you want corrected.
      >> Not ideal, but I think it will satisfy your needs.
      >> dmg

      Jim> The thing I remember most of PTStitchers color and exposure correction
      Jim> was how much it improved for my fisheye images when I used Crop or
      Jim> Select to eliminate the edge of the image circle. I have not tried
      Jim> PTBlender or looked at the algorithm but I believe two things would
      Jim> benefit the process. So both of these thing may already be implimented.

      Jim> First don't over correct Only correct each image 90 to 95 % of the way.
      Jim> That way a mistake is not propagated all the way done.

      This is a great idea. One way to do implement it is to have 2
      different projects, one "cropped" and one not. Then color correct the
      cropped, output curves and then use those curves with the
      non-cropped. Of course this is just a band-aid solution to do
      vignetting correction (see fulla for this).

      I guess ideally it would be nice to have a PTblender option that will
      use curves as input instead of estimating them itself.

      Jim> The second is when comparing the overlap of two image eliminate anything
      Jim> pieces for annalists in the overlap that is greatly different. This
      Jim> will help for moving items in the images.

      PTblender respects the mask to determine the pixels that overlap. I
      think masks can have "holes" in the middle of a region, so technically
      speaking you should be able to read the images and remove the "ghosts"
      and other information that is very different from one image to
      another. This, theoretically, should improve blending. It will be
      interesting to do it by removing most of the image, leaving only
      regions that are expected to be identical from one image to the next,
      like the sky. I'll be curious to know how it works if somebody tries

      Masks that are different than zero are considered to be part of the
      image. So you need to make sure you are using a totally black brush to
      paint over them

      Jim> --
      Jim> Jim Watters

      Jim> Yahoo ID: j1vvy ymsgr:sendIM?j1vvy
      Jim> jwatters @ photocreations . ca
      Jim> http://photocreations.ca

      Jim> -------------------------------------------------------------------------
      Jim> Take Surveys. Earn Cash. Influence the Future of IT
      Jim> Join SourceForge.net's Techsay panel and you'll get the chance to share your
      Jim> opinions on IT & business topics through brief surveys -- and earn cash
      Jim> http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
      Jim> _______________________________________________
      Jim> PanoTools-devel mailing list
      Jim> PanoTools-devel@...
      Jim> https://lists.sourceforge.net/lists/listinfo/panotools-devel

      Daniel M. German "As De Gaulle used to say:
      'Aim well, shoot fast
      Henri Cartier Bresson -> and get the hell out.'"
      dmg (at) uvic (dot) ca
      replace (at) with @ and (dot) with .
    Your message has been successfully submitted and would be delivered to recipients shortly.