Back to basics: Interpolation teaser
- From: "Pixel.to.life" <pixel.to.life@xxxxxxxxx>
- Date: Sat, 22 Sep 2007 18:11:36 -0000
Hi, All,
Here is a question on interpolation and resampling.
I take two images (a white line in the middle of a dark background
image, dimensions 128x128 pixels; and a circle of radius 31 pixel
units centered on a dark background image, dimensions 128x128 pixels).
http://medicalimagingscience.googlegroups.com/web/Original_Line.jpg
http://medicalimagingscience.googlegroups.com/web/Original_Circle.jpg
Now I resample them to double the image size along each dimension
using an appropriate interpolation method.
Here are the results:
http://medicalimagingscience.googlegroups.com/web/Resampled_Line.jpg
http://medicalimagingscience.googlegroups.com/web/Resampled_Circle.jpg
My questions are:
- what can be inferred about the nature of the interpolation method
employed given only the above information?
- on the resample circle image, the circle profile displays some 'saw
tooth' patterns. Is that an artifact/limitation of interpolation
scheme? Or it is expected?
- if you think there are artifacts in the resampled images, what could
be a possible solution or a better way to resample?
This is NOT a home work assignment. It is part of some simple tests on
a library I have been running:-)
Thanks.
Pixel.To.Life.
.
- Follow-Ups:
- Re: Back to basics: Interpolation teaser
- From: Martin Leese
- Re: Back to basics: Interpolation teaser
- From: Marco Al
- Re: Back to basics: Interpolation teaser
- Prev by Date: Re: Combining multiple photographic exposures into an HDR image
- Next by Date: Re: Back to basics: Interpolation teaser
- Previous by thread: fft2 different from FFTW 2d
- Next by thread: Re: Back to basics: Interpolation teaser
- Index(es):
Relevant Pages
|