GL_BLEND and Cuda

Accelerated Computing CUDA CUDA Programming and Performance
1

Hi everyone,

I just have n-images, which should be blended onto plane.
In opengl, it could be presented such as:

glBlendEqationEXt(gl_func_add_ext);
glBlendFunc(gl_one, gl_one);
glenable(gl_blend);
glBegin(GL_QUADS);

for(int i = 0; i < 5; i++)
{
      [_images[i] bindInput: GL_TEXTURE1_ARB];

      glMultiTexCoord4f(GL_TEXTURE1_ARB, -x, -y, z, 1);
      glVertex2f(0, 0);

      glMultiTexCoord4f(GL_TEXTURE1_ARB, -x, y, z, 1);
      glVertex2f(0, 1);

      glMultiTexCoord4f(GL_TEXTURE1_ARB, x, y, z, 1);
      glVertex2f(1, 1);

      glMultiTexCoord2f(GL_TEXTURE0_ARB, 1, 0);
      glMultiTexCoord4f(GL_TEXTURE1_ARB, x, -y, z, 1);
      glVertex2f(1, 0);
      glEnd();

      [_images[i] release: GL_TEXTURE1_ARB];
}
 glDisable(GL_BLEND);

Is there any way to implement blending by using cuda without opengl, namely I don’t want to visualize
blending.

Thanks a lot!