Curve Fitting and Curve Displaying
In Section 7.6 we discussed how one can define a curve on a discrete grid, and in Section 9.7 we discussed how to derive curves from skeletons. The representation of a curve as a sequence of pixels may be adequate for some applications, but for others we would like to have a mathematical expression for it. Such an expression may be far more compact than the discrete forms, as a comparison between Figure 9.8 and Table 9.1 shows. Finding a curve that passes through a set of given points is the problem of interpolation, while finding a curve that passes near a set of given points is the problem of approximation. We shall use the term curve fitting to refer collectively to both of them. The problem of curve displaying is also of interest: given a mathematical expression, identify the pixels that must be marked so that an image of the curve described by the expression is displayed. The problem is by no means trivial, even in the case of straight line displays (Section 7.6).
KeywordsHull Sine Dinate Editing
Unable to display preview. Download preview PDF.
- [10.CR]Crow, F. C. “The Use of Grayscale for Improved Raster Display of Vectors and Characters” SIGGRAPH’78, pp. 1–5.Google Scholar
- [10.KU]Kulpa, Z. “On the Properties of Discrete Circles, Rings, and Disks” CGIP, 10 (1979), pp. 348–365.Google Scholar
- [10.LL]Levy, H. and Lessman, F. Finite Difference Equations, New York: Macmillan, 1961.Google Scholar
- [10.WA]Warnock, J. F. “The Display of Characters Using Gray Level Sample Arrays,” SIGGRAPH’80, pp. 302–307.Google Scholar