Help for AUTOMATCH

PURPOSE
  AUTOMATCH is a program which automatically acquires tiepoints from
sets of images for the purpose of producing mosaics.
To obtain a more global view of the entire mosaicking process
you should look at the procedures in r3lib called:
MANUAL2.COM for Voyager, and MANUAL4.COM for Galileo. 
AUTOMATCH replaces MANMATCH in these procedures.

  Automatch is designed to generate tiepoints from sets of images which 
have been generated in a "burst" of frames where the scale & rotation
do not differ much between images. Provision is built into 
the program to compensate for differences between images by projecting
the correlation areas before correlation. This is limited however
to scale ranges of at most 2:1 and modest rotations which will
not confuse the overlap area algorithm. Automatch will NOT
work well on cases containing the limb.

EXECUTION

For VGR:
   AUTOMATCH  inp=(sedr.int,over.int,geoma.cor)  out=match.int                     project=VGR-1 filename=LISTOFFILES

For GLL:
   automatch inp=(jupsedr.int,jupover.int) out=jupmatch.int           filename=LISTOFFILES project=GLL

USAGE:
AUTOMATCH is an batch program which has in the SEDR.INT file
the FDS or SCET times of all of the frames to inspect along
with all the navigation for these images.. 
You must create this file using the program IBISNAV

AUTOMATCH can obtain the names (and thus the locations) of all the
input files in one of two ways:

(recommended)
1. By the user specifying the parameter FILENAME which points to
   a file created by a text editor and which contains the full
   file names of each input picture in the order of priority, one
   name per record. This is the same file used in IBISNAV.
   So, if the first 2 pictures are called scx1:[jjl059]1234567.img
   and mipldisk:[mipl.gll]venus.img the first two records of
   FILENAME would contain:

   scx1:[jjl059]1234567.img
   mipldisk:[mipl.gll]venus.img

(obsolete)
2. By building each file name by appending to the DIR keyword
   the integer contained in column 1 of the SEDR.INT file followed
   by .IMG. You must in this mode name the input pictures
   accordingly. 
   For Voyager, column 1 is the FDS times thus the frames must be
   named the same as their FDS times; ie: 7654321.img is the file
   name of the image with fds time 7654321. 
   For Galileo the FDS times are just 1,2,3,4.. so the file names are
   1.img   2.img etc.
   So, if DIR=scx1:[jjl059] and SEDR.INT has fds time 1234567 in column
   one then the built-up name would be:  scx1:[jjl059]1234567.img

The second input file 
contains the overlap frame pairs to present to the user.
This file is created by program MOSPLOT.

First input= SEDR.INT produced by program IBISNAV
Second input= OVER.INT produced by program MOSPLOT

Automatch will perform the following steps:

For each overlap pair:
1. Compute the overlap area in the right image by projecting the
   4 corners of the left image into the right image using the SEDR.
2. Select the correlation size to fit in the overlap area.
3. Select a row or column (whichever is the longest) through the
   overlap area and select a number of candidate tiepoint locations
   in the right image.
4. Compute the corresponding tiepoint positions in the left image
   using the SEDR.
5. Perform correlations.
6. Fit a 1st order polynomial to the tiepoints.
7. Iteratively reject all points (one at a time) exceeding a threshold
   pixel tolerance.
8. Select the 2 remaining points the furthest apart.
9. Place these in the output.

NOTE: 

The output file is the same as the file written by MANMATCH. So
if AUTOMATCH cannot process a certain number of overlap pairs you
can input the AUTOMATCH output into MANMATCH as the last input file
and go to the pairs in question to acquire points manually.

Be sure to inspect footprint plots of your mosaic generated by
MOSPLOT and to edit the overlap file with EDIBIS to delete 
extraneous & redundant overlaps to save time and later confusion.

At the moment automatch can accept up to 256 image pairs.

FILE STRUCTURE:
  The tiepoints file is an IBIS format tabular file. It consists
of 512 byte records where each column of data is written as
sequential records until exhausted. The next column begins at the
start of the next record etc. Record #1 contains the number of points
per column. All data is real*4 binary.

PARAMETERS: