Actually, you have some alternativs for the solution of your problem,
1. You can use Bitmap class to write your own resizing algorithm,
which can be bilinear or bicubic. For this you need to access points
of the image and modify the image. However this would rather be slow
both in running time and as setup cost.
2. As a second alternative you can install and run the Open Source
Computer Vision library, which is the best possible available image
processing library you can find. I was first developped by intel now
it's open source. you can find it from :
http://sourceforge.net/projects/opencvlibrary/ It comes
with lot's of image processing functions and it's really fast. If you
need the max. acceleration, you can use the Intel Performance
Primitives Library (IPP) which has, signal processing, matrix and
image processing sub libraries, and written in the most optimized way
possible. If you install it under Open CV, open CV will detect it and
use the ipp functions. you can see ipp as the lowest level of image
processing algorithms, and Open CV as a warpper for it.
Ps. You can fins Intel IPP easily with a search at Google or Intel's
web site.
