This preview shows page 1. Sign up to view the full content.
Unformatted text preview: so we
need to collect the incident light in a specular cone to
determine the specular reﬂection. !
• This is computationally infeasible. !
• We will return to specularities later, but for the moment
we will consider only diffuse radiosity. ! Graphics Lecture 13: Slide 17! Patching Problems "
• We need to divide our graphics scene into patches for
computing the radiosity. !
• For small polygons we can perhaps use the polygon
map, but for large polygons we need to subdivide them. !
• Since the emitted light will not be constant across a large
polygon we will see the subdivisions ! Graphics Lecture 13: Slide 18! Large Polygons "
• Each patch will have a different but constant illumination.!
• Thus we will see the patch boundaries unless either: !
– Patches project to (sub) pixel size or
– We smooth the results (eg by interpolation) ! Graphics Lecture 13: Slide 19! Form Factors"
• The form factors couple every pair of patches, determining
the proportion of radiated energy from one that strikes the
The form factors couple every pair of patches, determining the
proportion of radiated energy from one that strikes the other. 1
|Ai | Z Ai Z Aj cos cos
i j dAj dAi Graphics Lecture 13: Slide 20! . 16 / 29 Form Factors - the- deﬁnition
Form Factors the deﬁnition "
• The cos terms compute the projection of each patch in
The cos terms compute the projection of each patch in the
the direction to the other. !
direction to the other.
• If the patches are in the same plane, facing the same
Ifway, there areno the same plane, facing the same way, thereother
the patches is in coupling. If they directly face each is
they are If they directly face each
no coupling.maximally coupled. ! other they are maximally
|Ai | Z Ai Z cos Aj cos
i j dAj dAi Graphics Lecture 13: Slide 21! . 17 / 29 Form Factors - simplifying the computation"
Form Factors - simplifying the computa...
View Full Document
This document was uploaded on 03/26/2014.
- Spring '14