[ptx] Enblend Multiresolution Spline Blender

[Yili Zhao]

Hi,
  for the seam line, there are also some research papers:
  http://www.gisdevelopment.net/aars/acrs/1991/psq/ps016a.shtml
  o Milgram D.L., 1975. Computer methods for creating photomosaics. Trans IEEE on Computers, Vol. C-24, pp 1113-1119. 
  o Milgram D.L., 1977. Adaptive techniques for photomosaicing. Trans. IEEE on Computers, Vol. C-26, pp 1175-1186. 
  o Susanne Hummer, Miller, 1989. A Digital Mosaicing Algorithm allowing for an irregular joint line? Photo Engg. And R. Sensing Vol. 55, No. 1, Jan. 1989, pp 43-47.
  o Yang Shiren et al., 1989. Two dimensional Seam-point searching in Digital Image Mosaicing. Photo. Engg. And R. Sensing, Vol. 55, No. 1, Jan. 1989, pp 49-53. 
  o Peleg S., 1981. Elimination of Seam for Photomosaics. IEEE Conference on Pattern Recognition and Image Processing, pp 426-429. 
  o Marie Lise, Duplaquet. Building Large Image Mosaic with Invisible Seam line, Proceedings of SPIE, Vol 3387, Orlando, pages 369-377, 1998.

Best regards
Yili

======= 2004-03-10 04:49:58 Andrew wrote:=======

>I'm not convinced that it is enough to only know the distance to the 
>center of each image. This will cause you to pick a seam line which is 
>equidistant from the centers of two images. This is a straight line. For 
>irregular overlaps, this is not a good line. I think you need to know the 
>distance to the nearest image edge. Vigra has a distance transform - also 
>the Rosenfeld-Pfaltz method. I'll look into this some more.
>
>Andrew
>
>---------------------------------
>Andrew Mihal
>www-cad.eecs.berkeley.edu/~mihal
>mihal at eecs.berkeley.edu
>
>
>

= = = = = = = = = = = = = = = = = = = =