I did a brief search of the forums, and didn't find anything on collision detection that was 1) this general, or 2) had this kind of question
I don't want bounding box! every place i look all game tutorials use bounding box, and that is not accurate enough for what i want to do. here is something that is an example i drew:
this is a simple image i drew added to one image twice and rotated 45 deg the second time. the bounding box for the original image is 64 square, the rotated one became 66 square, which is to be expected when rotating an image. with a bounding box collision the yellow dot would be colliding with the image, in a simple game that wouldn't be a problem, but for a more advanced game it would be quite a problem, especially with a more complex character or other in-game being.
if we make the problem 3D it becomes much more complex, say for example we extrude the black shape out of the screen, and make the yellow dot a sphere. bounding box would again be a very imprecise.
most things that i can find on the internet are either language or library specific (eg. allegro uses bit masks for such stuff but that's only 2D) about things like this and i want more general stuff, not even code, just the math/ background behind it.
NOTE: i have not studied calculus yet! (so please try to make it simple :) )