Unformatted text preview: 348 MATHEMATICAL ELEMENTS FOR COMPUTER GRAPHICS In this case, the resulting BSpline curve passes through each data point, i.e., a
curve ﬁt is obtained. Although the continuity of the resulting curve is everywhere
Ck‘z, it may not be ‘smooth’, or ‘SWeet’ or ‘fair’. The ﬁtted curve may develop
unwanted wiggles or undulations. A fairer or smoother curve is obtained by specifying fewer deﬁning polygon
points than data points, i.e., 2 S k S n + 1 < j. Here, [ N ] is no longer square,
the problem is overspeciﬁed and can only be solved in a mean sense. Recalling
that a matrix times its transpose is always square (see Sec. 3—21), the deﬁning
polygon vertices for a B—spline curve that fairs or smooths the data are given» by [D]=[N][B]
[NlTlDl=lN]T[N][B] and [B]=[[N1T [Nil—llNlTlD] (5—117) Both of these techniques assume that the matrix [ N ] is known. Provided that
the order of the B—spline basis k, the number of deﬁning polygon points n + 1,
and the parameter value along the curve are known, then the basis functions
N§,k(tj) and hence the matrix [N ] can be obtained. Within the restrictions
2 S k S n + 1 S j, the order and number of polygon vertices are arbitrary. The parameter value tj for each data point is a measure of the data point’s
distance along the Bspline curve. One useful approximation for this parameter value uses the chord length between data points. Speciﬁcally, for 3' data points
the parameter value at the lth data point is t1=0 I
Z le _ «De—1 t: 5:2
1' 2 [De —‘ Ds—ll s=2 l>2 tmax The maximum parameter value tmax is usually taken as the maximum value
of the knot vector. Similar schemes are mentioned in Refs. (5—15) and (5—22). For an open uniform knot vector with multiplicity of the knot values at the
ends equal to k, a Bézier curve is obtained when n = k.
An example illustrates these techniques. Example 5—17 Bspline Curve Fit For the ﬁve data points D1[0 0 ], D2[1.5 2], D3[3 2.5 ], D4 [4.5 2], D5 [ 6 0] determine the thirdorder (k = 3) deﬁning polygons having ﬁve and four polygon vertices that generate a Bspline curve ‘through’ the data points. Use the chord length approximation for the parameter values along
. the B—spline curve corresponding to the data points. ...
View
Full Document
 Spring '10
 Dr.KaimingYU

Click to edit the document details