.help msisinit Dec98 "Image Interpolator Package" .ih NAME msisinit -- initialize the sequential interpolant descriptor using user parameters .ih SYNOPSIS include msisinit (msi, interp_type, nsinc, nxincr, nyincr, pixfrac1, pixfrac2, badval) .nf pointer msi #O interpolant descriptor int interp_type #I interpolant type int nsinc #I sinc interpolant width in pixels int nxincr,nyincr #I sinc look-up table resolution real pixfrac1,pixfrac2 #I sinc or drizzle pixel fractions real badval #I drizzle undefined pixel value .fi .ih ARGUMENTS .ls msi Pointer to sequential interpolant descriptor. .le .ls interp_type Interpolant type. The options are II_BINEAREST, II_BILINEAR, II_BIPOLY3, II_BIPOLY5, II_BISPLINE3, II_BISINC, II_BILSINC, and II_BIDRIZZLE for the nearest neighbour, linear, 3rd order polynomial, 5th order polynomial, cubic spline, sinc, look-up table sinc, and drizzle interpolants respectively. The interpolant type definitions are found in the package header file math/iminterp.h. .le .ls nsinc The sinc and look-up table sinc interpolant width in pixels. Nsinc is rounded up internally to the nearest odd number. .le .ls nxincr, nyincr The look-up table sinc resolution in x and y in number of entries. Nxincr = 10 implies a pixel resolution of 0.1 pixels in x, nxincr = 20 a pixel resolution of 0.05 pixels in x, etc. The default value of nxincr and nyincr are 20 and 20 .le .ls pixfrac1, pixfrac2 The look-up table sinc fractional pixel shifts in x and y if nincr = 1 in which case -0.5 <= rparam[1/2] <= 0.5 , or the drizzle pixel fractions in which case 0.0 <= rparam[1/2] <= 1.0. .le .ls badval The undefined pixel value for the drizzle interpolant. Pixels within the boundaries of the input image which do not overlap the input image pixels are assigned a value of badval. .le .ih DESCRIPTION The interpolant descriptor is allocated and initialized. The pointer msi is returned by MSISINIT. .ih NOTES MSIINIT or MSISINIT must be called before using any other MSI routines. .ih SEE ALSO msisinit, msifree