[gegl/samplers] comments
- From: Nicolas Robidoux <nrobidoux src gnome org>
- To: commits-list gnome org
- Cc:
- Subject: [gegl/samplers] comments
- Date: Thu, 30 Jun 2011 12:25:54 +0000 (UTC)
commit 2acbb8cd5ccb6d621b8becc140a4be45434384af
Author: Nicolas Robidoux <nicolas robidoux gmail com>
Date: Thu Jun 30 08:25:50 2011 -0400
comments
gegl/buffer/gegl-sampler-lohalo.c | 23 +++++++++++++++++------
1 files changed, 17 insertions(+), 6 deletions(-)
---
diff --git a/gegl/buffer/gegl-sampler-lohalo.c b/gegl/buffer/gegl-sampler-lohalo.c
index cec7293..4a2cfd6 100644
--- a/gegl/buffer/gegl-sampler-lohalo.c
+++ b/gegl/buffer/gegl-sampler-lohalo.c
@@ -40,6 +40,16 @@
*/
/*
+ * Reference:
+ *
+ * Nohalo subdivision (with bilinear instead of LBB "finish") is
+ * documented in Robidoux, N., Gong, M., Cupitt, J., Turcotte, A., and
+ * Martinez, K. CPU, SMP and GPU implementations of Nohalo level 1, a
+ * fast co-convex antialiasing image resampler. In Proceedings of
+ * C3S2E. 2009, 185-195.
+ */
+
+/*
* Credits and thanks:
*
* Jacobian adaptive resampling was developed by N. Robidoux and
@@ -2333,12 +2343,13 @@ gegl_sampler_lohalo_get ( GeglSampler* restrict self,
* Update using mipmap level 1 values.
*
* Possible future improvement: When the ellipse is
- * slanted, one could avoid many operations using
- * Anthony Thyssen's formulas for the bounding
- * parallelogram with horizontal top and bottom. When
- * both the magnification factors are the same, or
- * when there is no rotation, using these formulas
- * makes no difference.
+ * slanted, one could avoid many pixel value loads and
+ * operations with Anthony Thyssen's formulas for the
+ * ellipse bounding parallelogram with horizontal top
+ * and bottom. When both the magnification factors are
+ * the same, or when there is no rotation, using these
+ * formulas makes no difference. Reference:
+ * ImageMagick resample.c.
*/
{
gint i;
[
Date Prev][
Date Next] [
Thread Prev][
Thread Next]
[
Thread Index]
[
Date Index]
[
Author Index]