Occlusion

Convexity from an Object

I have just started reading about occlusion and immediately noticed the similarity with convexity in maths. The difference is that where all points inside a convex object are visible (straight line between them), for occlusion only one objects perspective is needed.

For example consider point based visibility in 2D in a rendering scene where for simplicity all the objects are solid. The final geometrical shape of the view every point can be seen from the one point.

The object need not only be a point. For a square view cell in 2D it is where any of the cells edges sees.

What can be seen can generally be transversed as a surface. In 2D the view or horizon is a line which is circular about the source. In 3D the horizon is a 2D surface about the point.