Using Opengl to transform images

I have a project where I need to be able to rotate, stretch, and skew images fairly quickly. The images will almost always be completely in memory (IE no file to deal with). So I figured "Hey, Graphics cards are good at the sort of stuff, why not use them?" The other benefit to this is that it would make the cross-platformness of the algorithm much easier to handle.

So here is what I would like my pipeline to look like. Get image data and put it in some OpenGL readable texture, Tell OpenGL to transform the texture, retrieve the transformed data from opengl and place it into some new image container (IE an HBITMAP or XImage).

So can you do something like that? or do I just have to rely on the cpu to process this stuff?

OpenGL is really a rendering API. You could use it for this but it seems like the wrong tool for the job. If you're running on linux or mac you could use the Image Magic tools (look specifically at mogrify).

I agree with Perspective that this does not sound like the sort of application that would benefit from OGL. I will add that D3D also will not help much. Perhaps a pixel shader or two could do some of what you want but I don't feel those are the right tool for this job.