Lecture 8: Mesh Representations and Geometry Processing (100)
ClaireLiu123
Is this used a similar idea to supersampling where this just breaks down each triangle and reforms a couple of small triangles inside?
LeslieTrue
@ClaireLiu123, Well, I think not that similar. The idea of supersampling is more like a low-pass filtering like doing convolution. Though it's, kind of a similar operation in implementation, i.e. dealing with each triangle and its neighbours, this simplification leads to quite opposite result with much more information loss.
jierui-cell
I am curious about whether there is a formal way of calculating the best number of triangles that provides the "best" quality with minimal cost.
Is this used a similar idea to supersampling where this just breaks down each triangle and reforms a couple of small triangles inside?
@ClaireLiu123, Well, I think not that similar. The idea of supersampling is more like a low-pass filtering like doing convolution. Though it's, kind of a similar operation in implementation, i.e. dealing with each triangle and its neighbours, this simplification leads to quite opposite result with much more information loss.
I am curious about whether there is a formal way of calculating the best number of triangles that provides the "best" quality with minimal cost.