next up previous contents
Next: 4.3 D-NURBS Equations of Up: 4 Formulation of D-NURBS Previous: 4.1 Curves

4.2 Surfaces

A D-NURBS surface has a similar structure to the curve. Proceeding analogously from (2), we define

  equation175

Again, the control points and weights comprise the generalized coordinates and are assembled into vectors tex2html_wrap_inline1965 , tex2html_wrap_inline1967 , and tex2html_wrap_inline1921 . Two subscripts are now associated with the generalized coordinates, reflecting the surface parameters u and v. For concreteness, we order the components in these vectors such that the second subscript varies faster than the first, although this convention does not affect the derived results.

As before, we can write tex2html_wrap_inline1975 instead of tex2html_wrap_inline1977 . By analogy to tex2html_wrap_inline1925 in (4) and (6), we obtain for the D-NURBS surface

  equation191

  equation194

However, the contents of the Jacobian tex2html_wrap_inline1931 differ from those in the curve case. To arrive at an explicit expression for tex2html_wrap_inline1931 , let tex2html_wrap_inline1985 , for tex2html_wrap_inline1987 , and tex2html_wrap_inline1989 , be a tex2html_wrap_inline1947 diagonal matrix whose entries are

displaymath1959

and let the 3-vector

displaymath1960

As before, the tex2html_wrap_inline1993 and tex2html_wrap_inline1995 are assembled into tex2html_wrap_inline1951 and tex2html_wrap_inline1955 , respectively. Hence,

displaymath1961

Note that tex2html_wrap_inline1931 is now a tex2html_wrap_inline2003 matrix.



Demetri Terzopoulos | Source Reference