Change image using OpenGL?

Question

I have an XRay imaging device. Due to some technical limitations, the detector is made of a heterogeneous pixel size and several inclined and partially overlapping plates. Image is distorted. The geometry of the detector is known exactly.

I need a function that converts these distorted images into a flat image with a uniform pixel size. I already did this with the CPU, but I would like to try using GPU in portable mode with OpenGL.

I have no experience with OpenGL programming, and most of the information I could find on the Internet was useless for this use. How do I proceed? How should I do it?

The image size is 560x860 pixels, and we have a batch of 720 images for processing. I'm on Ubuntu.

Answer 1

You may find this tutorial useful (it's a bit old, but note that it contains some OpenGL 2.x GLSL after the Cg section). I do not believe that GLSL has shortcuts for image processing, if that is what you are looking for ... you need to understand many aspects of 3D rasterization and historical baggage in order to use it efficiently, although once you do this there is a structure for inputs and outputs that you can forget about this and easily bypass your own algorithms in the shader code.

( Direct3D-, CUDA), , , CUDA/OpenCL. . , API API .

Answer 2

OpenGL . , , , OpenCL. - , , .

Answer 3

OpenGL CUDA GPGPU.

C ++, CUDA , , , .

Answer 4

OpengGL, , , , . , , CPU .

, , , , , ( ) CUDA, -, .

Answer 5

: " GPU ?"

, , , , < x, y > :

, x by & frac12;, y 1,2, :

rotation, shear ..

, -, , , , . . ( .., , , , .)

, , , , , , SIMD . , CUDA/GL/ .. , , , , , ..

ad-hoc- , .