INTL2B - bilinear interpolation of a 2-D array of bytes

intl2b		bilinear interpolation of a 2-D array of bytes

Function Prototype:
void intl2b (int nxin, float dxin, float fxin,
	int nyin, float dyin, float fyin, unsigned char *zin,
	int nxout, float dxout, float fxout,
	int nyout, float dyout, float fyout, unsigned char *zout);

Input:
nxin		number of x samples input (fast dimension of zin)
dxin		x sampling interval input
fxin		first x sample input
nyin		number of y samples input (slow dimension of zin)
dyin		y sampling interval input
fyin		first y sample input
zin		array[nyin][nxin] of input samples (see notes)
nxout		number of x samples output (fast dimension of zout)
dxout		x sampling interval output
fxout		first x sample output
nyout		number of y samples output (slow dimension of zout)
dyout		y sampling interval output
fyout		first y sample output

Output:
zout		array[nyout][nxout] of output samples (see notes)

Notes:
The arrays zin and zout must passed as pointers to the first element of
a two-dimensional contiguous array of unsigned char values.

Constant extrapolation of zin is used to compute zout for
output x and y outside the range of input x and y.
 
For efficiency, this function builds a table of interpolation
coefficents pre-multiplied by byte values.  To keep the table
reasonably small, the interpolation does not distinguish
between x and y values that differ by less than dxin/ICMAX
and dyin/ICMAX, respectively, where ICMAX is a parameter
defined above.

Author:  Dave Hale, Colorado School of Mines, c. 1989-1991.