[Bf-committers] Heckbert's original method of clamping the EWA (Elliptical Weighted Averaging) ellipse is not the best

Nicolas Robidoux nicolas.robidoux at gmail.com
Mon Feb 8 12:53:57 CET 2016


A better approach, originally suggested by Andreas Gustaffson in his 1993
Masters thesis, is used by ImageMagick and the GEGL (GEneric Graphic
Library). It relies, basically, on polar decomposition (by stretching the
minor and major axes of the ellipse if too short, keeping it's orientation
intact).

http://www.imagemagick.org/api/MagickCore/resample_8c_source.html
https://github.com/GNOME/gegl/blob/master/gegl/buffer/gegl-sampler-lohalo.c
https://github.com/GNOME/gegl/blob/master/gegl/buffer/gegl-sampler-nohalo.c


More information about the Bf-committers mailing list