Help for GTCROP

PURPOSE

     gtcrop cuts an image by a polygon, leaving the inside
     at the input image value and the outside all zero.  The
     polygon is stored in two columns of an IBIS table and
     is in (line,samp) coordinate of the image.  If you have
     geographic coordinates, use gtproj and/or pixmap to get
     the coordinates into image (line,samp).
     
     The user can choose whether the output is the same size
     as the input, or is reduced in size to barely contain 
     the polygon.  The latter case has two features:
     
     1.  A geotiff label is updated for the reduced image.
     2.  The starting (line,samp) of the reduced image in
         the original image can be output to TCL variables.
TAE COMMAND LINE FORMAT

     gtcrop INP=(PIC,INT) OUT=CUTPIC COLS=(XCOL,YCOL) 'COVERINP

     where

     PIC                 is an image.
     INT                 is    an   IBIS   interface    file 
                         containing the polygon that cuts the
                         image.
     CUTPIC              is  a copy of PIC with pixels outside 
                         of the polygon changed to zero.
     COLS                are the columns containing the polygon
                         in the IBIS table
     'COVERINP           causes CUTPIC to have the same size and
                         location as PIC, see other case below

     gtcrop INP=(PIC,INT) OUT=CUTPIC COLS=(XCOL,YCOL) 'COVERPOL

     where

     PIC                 is an image.
     INT                 is    an   IBIS   interface    file 
                         containing the polygon that cuts the
                         image.
     CUTPIC              is  a subimage of PIC with pixels outside 
                         of the polygon changed to zero.
     COLS                are the columns containing the polygon
                         in the IBIS table
     'COVERPOL           causes CUTPIC to have a reduced size, just
                         large enough to contain the intersection of
                         the poly with PIC
OPERATION
The pixel center is used to determine whether a pixel is
in or out.

"Inside" is determined in a primitive fashion by
determining the number of crossings (of the polygon)
by the infinite horizontal line containing the pixel.
The polygon is not checked for closedness, so a u-shaped
polygon will cut an image as well.  Closed polygons
(last point equals the first point) are recommended.
Difficult cases of polygon corners nearly exactly on
a line of pixels are handled by varying the polygon
by 1.e-8 of a pixel usually in the larger direction.

The polygon should be in the cyclic order (line,sample,
line,sample,...) in the columns specified by the COLS
parameter.


WRITTEN BY:                   Al Zobrist         18 JUL 00
COGNIZANT PROGRAMMER:         B. McGuffie
Revisions:

  2008-01-11 WLB Switched to USES_ANSI_C AND LIB_CARTO; misc cleanup  
  2015-10-15 WLB Migrated to mipl