.help revisions Jan97 images.imgeom
.nf

===============================
Package Reorganization
===============================

pkg/images/imarith/t_imsum.x
pkg/images/imarith/t_imcombine.x
pkg/images/doc/imsum.hlp
pkg/images/doc/imcombine.hlp
    Provided options for USHORT data.  (12/10/96, Valdes)

pkg/images/imarith/icsetout.x
pkg/images/doc/imcombine.hlp
    A new option for computing offsets from the image WCS has been added.
    (11/30/96, Valdes)

pkg/images/imarith/t_imcombine.x
pkg/images/imarith/icombine.gx
    Changed the error checking to catch additional errors relating to too
    many files.  (11/12/96, Valdes)

pkg/images/imarith/icsort.gx
    There was an error in the ic_2sort routine when there are exactly
    three images that one of the explicit cases did not properly keep
    the image identifications.  See buglog 344.  (8/1/96, Valdes)

pkg/images/filters/median.x
    The routine mde_yefilter was being called with the wrong number of
    arguments.
    (7/18/96, Davis)

pkg/images/imarith/t_imcombine.x
pkg/images/imarith/icombine.gx
pkg/images/imarith/icimstack.x +
pkg/images/imarith/iclog.x
pkg/images/imarith/mkpkg
pkg/images/doc/imcombine.hlp
    The limit on the maximum number of images that can be combined, set by
    the maximum number of logical file descriptors, has been removed.  If
    the condition of too many files is detected the task now automatically
    stacks all the images in a temporary image and then combines them with
    the project option.
    (5/14/96, Valdes)

pkg/images/geometry/xregister/rgxfit.x
    Changed several Memr[] references to Memi[] in the rg_fit routine.
    This bug was causing a floating point error in the xregister task
    on the Dec Alpha if the coords file was defined, and could potentially
    cause problems on other machines.
    (Davis, April 3, 1996)

pkg/images/geometry/t_geotran.x
pkg/images/geometry/geograph.x
pkg/images/doc/geomap.hlp
     Corrected the definition of skew in the routines which compute a geometric
     interpretation of the 6-coefficient fit, which compute the coefficients
     from the geometric parameters, and in the relevant help pages.
     (2/19/96, Davis)

pkg/images/median.par
pkg/images/rmedian.par
pkg/images/mode.par
pkg/images/rmode.par
pkg/images/fmedian.par
pkg/images/frmedian.par
pkg/images/fmode.par
pkg/images/frmode.par
pkg/images/doc/median.hlp
pkg/images/doc/rmedian.hlp
pkg/images/doc/mode.hlp
pkg/images/doc/rmode.hlp
pkg/images/doc/fmedian.hlp
pkg/images/doc/frmedian.hlp
pkg/images/doc/fmode.hlp
pkg/images/doc/frmode.hlp
pkg/images/filters/t_median.x
pkg/images/filters/t_rmedian.x
pkg/images/filters/t_mode.x
pkg/images/filters/t_rmode.x
pkg/images/filters/t_fmedian.x
pkg/images/filters/t_frmedian.x
pkg/images/filters/t_fmode.x
pkg/images/filters/t_frmode.x
    Added a verbose parameter to the median, rmedian, mode, rmode, fmedian,
    frmedian, fmode, and frmode tasks. (11/27/95, Davis)

pkg/images/geometry/doc/geotran.hlp
    Fixed an error in the help page for geotran. The default values for
    the xscale and yscale parameters were incorrectly listed as INDEF,
    INDEF instead of 1.0, 1.0. (11/14/95, Davis)

pkg/images/imarith/icpclip.gx
    Fixed a bug where a variable was improperly used for two different
    purposes causing the algorithm to fail (bug 316).  (10/19/95, Valdes)

pkg/images/doc/imcombine.hlp
    Clarified a point about how the sigma is calculated with the SIGCLIP
    option.  (10/11/95, Valdes)

pkg/images/imarith/icombine.gx
    To deal with the case of readnoise=0. and image data which has points with
    negative mean or median and very small minimum readnoise is set
    internally to avoid computing a zero sigma and dividing by it.  This
    applies to the noise model rejection options.  (8/11/95, Valdes)

pkg/images/frmedian.hlp
pkg/images/frmode.hlp
pkg/images/rmedian.hlp
pkg/images/rmode.hlp
pkg/images/frmedian.par
pkg/images/frmode.par
pkg/images/rmedian.par
pkg/images/rmode.par
pkg/images/filters/frmedian.h
pkg/images/filters/frmode.h
pkg/images/filters/rmedian.h
pkg/images/filters/rmode.h
pkg/images/filters/t_frmedian.x
pkg/images/filters/t_frmode.x
pkg/images/filters/t_rmedian.x
pkg/images/filters/t_rmode.x
pkg/images/filters/frmedian.x
pkg/images/filters/frmode.x
pkg/images/filters/rmedian.x
pkg/images/filters/rmode.x
pkg/images/filters/med_utils.x
    Added new ring median and modal filtering tasks frmedian, rmedian,
    frmode, and rmode to the images package.
    (6/20/95, Davis)

pkg/images/fmedian.hlp
pkg/images/fmode.hlp
pkg/images/median.hlp
pkg/images/mode.hlp
pkg/images/fmedian.par
pkg/images/fmode.par
pkg/images/median.par
pkg/images/mode.par
pkg/images/filters/fmedian.h
pkg/images/filters/fmode.h
pkg/images/filters/median.h
pkg/images/filters/mode.h
pkg/images/filters/t_fmedian.x
pkg/images/filters/t_fmode.x
pkg/images/filters/t_median.x
pkg/images/filters/t_mode.x
pkg/images/filters/fmedian.x
pkg/images/filters/fmode.x
pkg/images/filters/median.x
pkg/images/filters/mode.x
pkg/images/filters/fmd_buf.x
pkg/images/filters/fmd_hist.x
pkg/images/filters/fmd_maxmin.x
pkg/images/filters/med_buf.x
pkg/images/filters/med_sort.x
    Added minimum and maximum good data parameters to the fmedian, fmode,
    median, and mode filtering tasks.  Removed the 64X64 kernel size limit
    in the median and mode tasks.  Replaced the common blocks with structures
    and .h files.
    (6/20/95, Davis)

pkg/images/geometry/t_geotran.x
pkg/images/geometry/geotran.x
pkg/images/geometry/geotimtran.x
    Fixed a bug in the buffering of the x and y coordinate surface interpolants
    which can cause a memory corruption error if, nthe nxsample or nysample
    parameters are > 1, and the nxblock or nyblock parameters are less
    than the x and y dimensions of the input image. Took the opportunity
    to clean up the code.
    (6/13/95, Davis)

=======
V2.10.4
=======

pkg/images/geometry/t_geomap.x
    Corrected a harmless typo in the code which determines the minimum
    and maximum x values and improved the precision of the test when the
    input is double precision.
    (4/18/95, Davis)

pkg/images/doc/fit1d.hlp
    Added a description of the interactive parameter to the fit1d help page.
    (4/17/95, Davis)

pkg/images/imarith/t_imcombine.x
    If an error occurs while opening an input image header the error
    recovery will close all open images and then propagate the error.
    For the case of running out of file descriptors with STF format
    images this will allow the error message to be printed rather
    than the error code.  (4/3/95, Valdes)

pkg/images/geometry/xregister/t_xregister.x
    Added a test on the status code returned from the fitting routine so
    the xregister tasks does not go ahead and write an output image when
    the user quits the task in in interactive mode.
    (3/31/95, Davis)

pkg/images/imarith/icscale.x
pkg/images/doc/imcombine.hlp
    The behavior of the weights when using both multiplicative and zero
    point scaling was incorrect; the zero levels have to account for
    the scaling.  (3/27/95, Valdes)

pkg/images/geometry/xregister/rgxtools.x
    Changed some amovr and amovi calls to amovkr and amovki calls.
    (3/15/95, Davis)

pkg/images/geometry/t_imshift.x
pkg/images/geometry/t_magnify.x
pkg/images/geometry/geotran.x
pkg/images/geometry/xregister/rgximshift.x
    The buffering margins set for the bicubic spline interpolants were
    increased to improve the flux conservation properties of the interpolant
    in cases where the data is undersampled. (12/6/94, Davis)

pkg/images/xregister/rgxbckgrd.x
    In several places the construct array[1++nx-wborder] was being used
    instead of array[1+nx-wborder]. Apparently caused by a typo which
    propagated through the code, the Sun compilers did not catch this, but
    the IBM/RISC6000 compilers did. (11/16/94, Davis)


pkg/images/xregister.par
pkg/images/doc/xregister.hlp
pkg/images/geometry/xregister/t_xregister.x
pkg/images/geometry/xregister/rgxcorr.x
pkg/images/geometry/xregister/rgxicorr.x
pkg/images/geometry/xregister/rgxcolon.x
pkg/images/geometry/xregister/rgxdbio.x
    The xregister task was modified to to write the output shifts file
    in either text database format (the current default)  or in simple text
    format. The change was made so that the output of xregister could
    both be edited more easily by the user and be used directly with the
    imshift task. (11/11/94, Davis)

pkg/images/imfit/fit1d.x
    A Memc in the ratio output option was incorrectly used instead of Memr
    when the bug fix of 11/16/93 was made.  (10/14/94, Valdes)

pkg/images/geometry/xregister/rgxcorr.x
    The procedure rg_xlaplace was being incorrectly declared as an integer
    procedure.
    (8/1/94, Davis)

pkg/images/geometry/xregister/rgxregions.x
    The routine strncmp was being called (with a missing number of characters
    argument) instead of strcmp. This was causing a bus error under solaris
    but not sun os whenever the user set regions to "grid ...". (7/27/94 LED)

pkg/images/tv/imexaine/ierimexam.x
    The Gaussian fitting can return a negative sigma**2 which would cause
    an FPE when the square root is taken.  This will only occur when
    there is no reasonable signal.  The results of the gaussian fitting
    are now set to INDEF if this unphysical result occurs.  (7/7/94, Valdes)

pkg/images/geometry/geofit.x
    A routine expecting two char arrays was being passed two real arrays
    instead resulting in a segmentation violation if calctype=real
    and reject > 0.
    (6/21/94, Davis)

pkg/images/imarith/t_imarith.x
    IMARITH now deletes the CCDMEAN keyword if present.  (6/21/94, Valdes)

pkg/images/imarith/icaclip.gx
pkg/images/imarith/iccclip.gx
pkg/images/imarith/icpclip.gx
pkg/images/imarith/icsclip.gx
    1.	The restoration of deleted pixels to satisfy the nkeep parameter
	was being done inside the iteration loop causing the possiblity
	of a non-terminating loop; i.e. pixels are rejected, they are
	restored, and the number left then does not statisfy the termination
	condition.  The restoration step was moved following the iterative
	rejection.
    2.	The restoration was also incorrectly when mclip=no and could
	lead to a segmentation violation.
    (6/13/94, Valdes)

pkg/images/geometry/xregister/rgxicorr.x
    The path names to the xregister task interactive help files was incorrect.
    (6/13/94, Davis)

pkg/images/imarith/iccclip.gx
pkg/images/imarith/icsclip.gx
    Found and fixed another typo bug.  (6/7/94, Valdes/Zhang)

pkg/images/imarith/icscale.x
    The sigma scaling flag, doscale1, would not be set in the case of
    a mean offset of zero though the scale factors could be different.
    (5/25/94, Valdes/Zhang)

pkg/images/imarith/icsclip.gx
    There was a missing line: l = Memi[mp1].  (5/25/94, Valdes/Zhang)

pkg/images/imarith/icaclip.gx
pkg/images/imarith/iccclip.gx
pkg/images/imarith/icpclip.gx
pkg/images/imarith/icsclip.gx
    The reordering step when a central median is used during rejection
    but the final combining is average was incorrect if the number
    of rejected low pixels was greater than the number of pixel
    number of pixels not rejected.  (5/25/94, Valdes)

pkg/images/geometry/t_geotran.x
    In cases where there was no input geomap database, geotran was
    unnecessarily  overiding the size of the input image requested by the
    user if the size of the image was bigger than the default output size
    (the size of the output image which would include all the input image
    pixels is no user shifts were applied).
    (5/10/94, Davis)

pkg/images/imarith/icscale.x
pkg/images/imarith/t_imcombine.x
    1.  There is now a warning error if the scale, zero, or weight type
	is unknown.
    2.  An sfree was being called before the allocated memory was finished
	being used.
    (5/2/94, Valdes)

pkg/images/tv/imexaine/ierimexam.x
    For some objects the moment analysis could fail producing a floating
    overflow error in imexamine, because the code was trying to use
    INDEF as the initial value of the object fwhm.  Changed the gaussian
    fitting code to use a fraction of the fitting radius as the initial value
    for the fitted full-width half-maximum in cases where the moment analysis
    cannot compute an initial value.
    (4/15/94 LED)

pkg/images/imarith/iclog.x
    Changed the mean, median, mode, and zero formats from 6g to 7.5g to
    insure 5 significant digits regardless of signs and decimal points.
    (4/13/94, Valdes)

pkg/images/doc/imcombine.hlp
    Tried again to clarify the scaling as multiplicative and the offseting
    as additive for file input and for log output.  (3/22/94, Valdes)

pkg/images/imarith/iacclip.gx
pkg/images/imarith/iccclip.gx
pkg/images/imarith/iscclip.gx
    The image sigma was incorrectly computed when an offset scaling is used.
    (3/8/94, Valdes)

pkg/images/doc/imcombine.hlp
    The MINMAX example confused low and high.  (3/7/94, Valdes)

pkg/images/geometry/t_geomap.x
pkg/images/geometry/geofit.x
pkg/images/geometry/geograph.x
    Fixed a bug in the geomap code which caused the linear portion of the transformation
    to be computed incorrectly if the x and y fits had a different functional form.
    (12/29/93, Davis)

pkg/images/imarith/t_imcombine.x
pkg/images/imcombine.par
pkg/images/do/imcombine.hlp
    The output pixel datatypes now include unsigned short integer.
    (12/4/93, Valdes)

pkg/images/doc/imcombine.hlp
    Fixed an error in the example of offseting.  (11/23/93, Valdes)

pkg/images/imfit/fit1d.x
    When doing operations in place the input and output buffers are the
    same and the difference and ratio operations assumed they were not
    causing the final results to be wrong.  (11/16/93, Valdes)

pkg/images/imarith/t_imarith.x
pkg/images/doc/imarith.hlp
    If no calculation type is specified then it will be at least real
    for a division.  Since the output pixel type defaults to the
    calculation type if not specified this will also result in a
    real output if dividing two integer images.  (11/12/93, Valdes)

pkg/images/imarith/icgrow.gx
pkg/images/imarith/icpclip.gx
pkg/images/imarith/icsclip.gx
pkg/images/imarith/icaclip.gx
pkg/images/imarith/iccclip.gx
pkg/images/imarith/t_imcombine.x
pkg/images/doc/imcombine.hlp
    If there were fewer initial pixels than specified by nkeep then the
    task would attempt to add garbage data to achieve nkeep pixels.  This
    could occur when using offsets, bad pixel masks, or thresholds.  The
    code was changed to check against the initial number of pixels rather
    than the number of images.  Also a negative nkeep is no longer
    converted to a positive value based on the number of images.  Instead
    it specifies the maximum number of pixels to reject from the initial
    set of pixels.  (11/8/93, Valdes)

=======
V2.10.2
=======

pkg/images/imarith/icsetout.x
    Added MWCS calls to update the axis mapping when using the project
    option in IMCOMBINE.  (10/8/93, Valdes)

pkg$images/imarith/icscale.x
pkg$images/doc/imcombine.hlp
    The help indicated that user input scale or zero level factors
    by an @file or keyword are multiplicative and additive while the
    task was using then as divisive and subtractive.  This was
    corrected to agree with the intend of the documentation.
    Also the factors are no longer normalized.  (9/24/93, Valdes)

pkg$images/imarith/icsetout.x
    The case in which absolute offsets are specified but the offsets are
    all the same did not work correctly.  (9/24/93, Valdes)

pkg$images/imfit/imsurfit.h
pkg$images/imfit/t_imsurfit.x
pkg$images/imfit/imsurfit.x
pkg$images/lib/ranges.x
    Fixed two bugs in the imsurfit task bad pixel rejection code. For low
    k-sigma rejections factors the bad pixel list could overflow resulting
    in a segmentation violation or a hung task. Overlapping ranges were
    not being decoded into a bad pixel list properly resulting in 
    oscillating bad pixel rejection behavior where certain groups of
    bad pixels were alternately being included and excluded from the fit.
    Both bugs are fixed in iraf 2.10.3
    (9/21/93, Davis)

pkg$images/doc/imcombine.hlp
    Clarified how bad pixel masks work with the "project" option.
    (9/13/93, Valdes)

pkg$images/imfit/fit1d.x
    When the input and output images are the same there was an typo error
    such that the output was opened separately but then never unmapped
    resulting in the end of the image not being updated.  (8/6/93, Valdes)

pkg$images/imarith/t_imcombine.x
    The algorithm for making sure there are enough file descriptors failed
    to account for the need to reopen the output image header for an
    update.  Thus when the number of input images + output images + logfile
    was exactly 60 the task would fail.  The update occurs when the output
    image is unmapped so the solution was to close the input images first
    except for the first image whose pointer is used in the new copy of the
    output image.  (8/4/93, Valdes)

pkg$images/filters/t_mode.x
pkg$images/filters/t_median.x
    Fixed a bug in the error trapping code in the median and mode tasks.
    The call to eprintf contained an extra invalid error code agument.
    (7/28/93, Davis)

pkg$images/geometry/geomap.par
pkg$images/geometry/t_geomap.x
pkg$images/geometry/geogmap.x
pkg$images/geometry/geofit.x
    Fixed a bug in the error handling code in geomap which was producing
    a segmentation violation on exit if the user's coordinate list
    had fewer than 3 data points. Also improved the error messages
    presented to the user in both interactive and non-interactive mode.
    (7/7/93, Davis)

pkg$images/imarith/icgdata.gx
    There was an indexing error in setting up the ID array when using
    the grow option.  This caused the CRREJECT/CCDCLIP algorithm to
    fail with a floating divide by zero error when there were non-zero
    shifts.  (5/26/93, Valdes)

pkg$images/imarith/icmedian.gx
    The median calculation is now done so that the original input data
    is not lost.  This slightly greater inefficiency is required so
    that an output sigma image may be computed if desired.  (5/10/93, Valdes)

pkg$images/geometry/t_imshift.x
    Added support for type ushort to the imshift task in cases where the
    pixel shifts are integral.
    (5/8/93, Davis)

pkg$images/doc/rotate.hlp
    Fixed a bug in the rotate task help page which implied that automatic
    image size computation would occur if ncols or nlines were set no 0
    instead of ncols and nlines.
    (4/17/93, Davis)

pkg$images/imarith/imcombine.gx
    There was no error checking when writing to the output image.  If
    an error occurred (the example being when an imaccessible imdir was
    set) obscure messages would result.  Errchks were added.
    (4/16/93, Valdes)

pkg$images/doc/gauss.hlp
    Fixed 2 sign errors in the equations in the documentation describing
    the elliptical gaussian fucntion.
    (4/13/92, Davis)

pkg/images/imutil/t_imslice.x
    Removed an error check in the imslice task, which was preventing it from
    being used to reduce the dimensionality of images where the length of 
    the slice dimension is 1.0.
    (2/16/83, Davis)

pkg/images/filters/fmedian.x
    The fmedian task was printing debugging information under iraf 2.10.2.
    (1/25/93, Davis)

pkg/images/imarith/icaclip.gx
pkg/images/imarith/iccclip.gx
pkg/images/imarith/icpclip.gx
pkg/images/imarith/icsclip.gx
    When using mclip=yes and when more pixels are rejected than allowed by
    the nkeep parameter there was a subtle bug in how the pixels are added
    back which can result in a segmentation violation.
	if (nh == n2)  ==>  if (nh == n[i])
    (1/20/93, Valdes)


=======
V2.10.1
=======

pkg/images/imarith/t_imcombine.x
pkg/images/imarith/icaclip.gx
pkg/images/imarith/iccclip.gx
pkg/images/imarith/icgrow.gx
pkg/images/imarith/iclog.x
pkg/images/imarith/icombine.com
pkg/images/imarith/icombine.gx
pkg/images/imarith/icombine.h
pkg/images/imarith/icpclip.gx
pkg/images/imarith/icscale.x
pkg/images/imarith/icsclip.gx
pkg/images/imarith/icsetout.x
pkg/images/imcombine.par
pkg/images/doc/combine.hlp
    The weighting was changed from using the square root of the exposure time
    or image statistics to using the values directly.  This corresponds
    to variance weighting.  Other options for specifying the scaling and
    weighting factors were added; namely from a file or from a different
    image header keyword.  The \fInkeep\fR parameter was added to allow
    controlling the maximum number of pixels to be rejected by the clipping
    algorithms.  The \fIsnoise\fR parameter was added to include a sensitivity
    or scale noise component to the noise model.  Errors will now delete
    the output image.
    (9/30/92, Valdes)

pkg/images/imutil/imcopy.x
    Added a call to flush after the status line printout so that the output
    will appear immediately. (8/19/92, Davis)

pkg/images/filters/mkpkg
pkg/images/filters/t_fmedian.x
pkg/images/filters/fmedian.x
pkg/images/filters/fmd_buf.x
pkg/images/filters/fmd_maxmin.x
    The fmedian task could crash with a segmentation violation if mapping
    was turned off (hmin = zmin and hmax = zmax) and the input image
    contained data outside the range defined by zmin and zmax. (8/18/92, Davis)

pkg/images/imarith/icaclip.gx
pkg/images/imarith/iccclip.gx
pkg/images/imarith/icpclip.gx
pkg/images/imarith/icsclip.gx
    There was a very unlikely possibility that if all the input pixels had
    exactly the same number of rejected pixels the weighted average would
    be done incorrectly because the dflag would not be set.  (8/11/92, Valdes)

pkg/images/imarith/icmm.gx
    This procedure failed to set the dflag resulting in the weighted average
    being computed in correctly.  (8/11/92, Valdes)

pkg/images/imfit/fit1d.x
    At some point changes were made but not documented dealing with image
    sections on the input/output.  The changes seem to have left off the
    final step of opening the output image using the appropriate image
    sections.  Because of this it is an error to use an image section
    on an input image when the output image is different; i.e.
    
	cl> fit1d dev$pix[200:400,*] junk
	
    This has now been fixed.  (8/10/92, Valdes)

pkg/images/imarith/icscales.x
    The zero levels were incorrectly scaled twice.  (8/10/92, Valdes)

pkg/images/imarith/icstat.gx
    Contained the statement
	nv = max (1., (Memi[v2+i] - Memi[v1+i]) / Memi[dv+i] + 1)
    which is max(real,int).  Changed the 1. to a 1.  (8/10/92, Valdes)

pkg$images/imarith/icaclip.gx
pkg$images/imarith/iccclip.gx
pkg$images/imarith/icsclip.gx
    These files contained multiple cases (ten or so) of constructs such as
    "max (1., ...)" or "max (0., ...)" where the ... could be either real
    or double.   In the double cases the DEC compiler complained about a
    type mismatch since 1. is real.  (8/10/92, Valdes)

pkg$images/imfit/t_imsurfit.x
    Fixed a bug in the section reading code. Imsurfit is supposed to switch
    the order of the section delimiters in x and y if x2 < x1 or y2 < 1.
    Unfortunately the y test was actually "if (y2 < x1)" instead of 
    "if (y2 < y1)". Whether or not the code actually works correctly
    depends on the value of x1 relative to y2. This bug was not present
    in 2.9.1 but is present in subsequent releases. (7/30/92 LED)

=======
V2.10.1
=======

pkg$images/filters/t_gauss.x
    The case theta=90 and ratio > 0.0 but  < 1.0 was producing an incorrect
    convolution if bilinear=yes, because the major axis sigmas being 
    input along the x and y axes were sigma and ratio * sigma respectively
    instead of ratio * sigma and sigma in this case.

pkg$images/imutil/imcopy.x
    Modified imcopy to write its verbose output to STDOUT instead of
    STDERR. (6/24/92, Davis)

pkg$images/imarith/imcombine.gx
    The step where impl1$t is called to check if there is enough memory
    did not set the return buffer because the values are irrelevant for
    this check.  However, depending on history, this buffer could have
    arbitrary values and later when IMIO attempts to flush this buffer,
    at least in the case of image type coersion, cause arithmetic errors.
    The fix was to clear the returned buffers.  (4/27/92, Valdes)

pkg$images/imutil/t_imstack.x
    Modified the imslice task to read the old and write a new axis map.
    (4/23/92, Davis)

pkg$images/geometry/t_imslice.x
    Modified the imslice task to read the old and write a new axis map.
    (4/23/92, Davis)

pkg$images/geometry/t_blkavg.x
pkg$images/geometry/t_blkrep.x
    Modified the calls to mw_shift and mw_scale to explicitly set the
    number of logical axes instead of using the default of 0.
    (4/23/92, Davis)

pkg$images/geometry/t_imtrans.x
    Modified imtranspose so that it correctly picks up the axis map
    and writes it to the output image wcs. (4/23/92, Davis)

pkg$images/register.par
pkg$images/geotran.par
pkg$images/doc/register.hlp
pkg$images/doc/geotran.hlp
    Changed the default values of the parameters xscale and yscale in
    the register and geotran tasks from INDEF to 1.0 (4/23/92, Davis)

pkg$images/geometry/t_imtrans.x
pkg$images/doc/imtranspose.hlp
    Modified the imtranspose task so it does a true transpose of the
    axes instead of simply modifying the lterm. (4/8/92, Davis)

pkg$images/iminfo/listpixels.x
    Added the formats parameter for formatting the output pixel coordinates
    to the listpixels task. These formats take precedence over the formats
    stored in the WCS in the image header and the previous default format.
    (4/7/92, Davis)

pkg$images/imutil/t_imstack.x
    Added wcs support to the imstack task. (4/2/92, Davis)

pkg$images/iminfo/listpixels.x
    Modified listpixels so that it will work correctly  if the dimension
    of the wcs is less than the dimension of the image. (3/16/92, Davis)

pkg$images/geometry/t_geotran.x
    Modified the rotate, imlintran, register and geotran tasks wcs updating
    code to deal correclty with dimensionally reduced data. (3/16/92, Davis)

pkg$images/imarith/icalip.gx
pkg$images/imarith/icclip.gx
pkg$images/imarith/ipslip.gx
pkg$images/imarith/icslip.gx
pkg$images/imarith/icmedian.gx
    The median calculation with an even number of points for short data
    could overflow (addition of two short values) and be incorrect.
    (3/16/92, Valdes)

pkg$images/geometry/t_blkavg.x
pkg$images/geometry/t_blkrep.x
    1. Improved the precision of the blkavg task wcs updating code.
    2. Changed the blkrep task wcs updating code so that it is consistent
    with blkavg. This means that a blkrep command followed by a blkavg
    command or vice versa will return the original coordinate system
    to within machine precision. (3/16/92, Davis)

pkg$images/iminfo/listpixels.x
    Modified listpixels to print out an error if it could not open the
    wcs in the image. (3/15/92, Davis)

pkg$images/geometry/t_magnify.x
    Fixed a bug in the magnify task wcs updating code which was not
    working correctly for dimensionally reduced images. (3/15/92, Davis)

pkg$images/geometry/t_imtrans.x
    Fixed a bug in the imtranspose task wcs updating code which was not
    working correctly for dimensionally reduced images. (3/14/92, Davis)

pkg$images/imarith/icalip.gx
pkg$images/imarith/icclip.gx
pkg$images/imarith/icslip.gx
    There was a bug allowing the number of valid pixels counter to become
    negative.  Also there was a step which should not be done if the
    number of valid pixels is less than 1; i.e. all pixels rejected.
    A test was put in to skip this step.  (3/13/92, Valdes)

pkg$images/iminfo/t_imslice.x
pkg$images/doc/imslice.hlp
    Added wcs support to the imslice task.
    (3/12/92, Davis)

pkg$images/iminfo/t_imstat.x
    Fixed a bug in the code for computing the standard deviation, kurtosis,
    and skew, wherein precision was being lost because two of the intermediate
    variables in the computation were real instead of double precision.
    (3/10/92, Davis)

pkg$images/iminfo/listpixels.x
    1. Modified listpixels task to use the MWCS axis "format" attributes 
    if they are present in the image header.
    2. Added support for dimensionally reduced images, i.e.
    images which are sections of larger images and whose coordinate
    transformations depend on the reduced axes, to the listpixels task. 
    (3/6/92, Davis)

pkg$images/imarith/t_imcombine.x
pkg$images/imarith/icsetout.x
    Changed error messages to say IMCOMBINE instead of ICOMBINE.
    (3/2/92, Valdes)

pkg$images/imarith/iclog.x
    Added listing of read noise and gain.  (2/10/92, Valdes)

pkg$images/imarith/icscale.x
pkg$images/imarith/icpclip.gx
    1.  Datatype declaration for asumi was incorrect.
    2.  Reduced the minimum number of images allowed for PCLIP to 3.
    (1/7/92, Valdes)

pkg$images/imarith/icgrow.gx
    The first pixel to be checked was incorrectly set to 0 instead of 1
    resulting in a segvio when using the grow option.  (12/6/91, Valdes)

pkg$images/imarith/icgdata.gx
pkg$images/imarith/icscale.x
    Fixed datatype declaration errors found by SPPLINT. (11/22/91, Valdes)

pkg$images/iminfo/t_imstat.x
    Fixed a bug in the kurtosis computation found by ST.
    (Davis 10/11/91)

pkg$images/iminfo/t_imstat.x
pkg$images/doc/imstat.hlp
    Corrected a bug in the mode computation in imstatistics. The parabolic
    interpolation correction for computing the histogram peak was being
    applied in the wrong direction. Note that for dev$pix the wrong answer
    is actually closer to the expected answer than the correct answer
    due to binning effects. 
    (Davis 9/24/91)

pkg$images/filters/t_gauss.x
    The code which computes the gaussian kernel was producing a divide by
    zero error if ratio=0.0 and bilinear=yes (2.10 version only).
    (Davis 9/18/91)

pkg$images/doc/magnify.hlp
    Corrected a bug in the magnify help page.
    (Davis 9/18/91)

pkg$images/imarith/icsclip.gx
pkg$images/imarith/icaclip.gx
pkg$images/imarith/iccclip.gx
    There was a typo,  Memr[d[k]+k] --> Memr[d[j]+k].  (9/17/91, Valdes)

pkg$images/imarith/icstat.gx
pkg$images/imarith/icmask.x
    The offsets were used improperly in computing image statistics.
    (Valdes, 9/17/91)

pkg$images/geometry/t_imshift.x
    The shifts file pointer was not being correctly initialized to NULL 
    in the case where no shifts file was declared. When the task
    was invoked repeatedly from a script, this could result in an array being
    referenced, for which space had not been previously allocated.
    (Davis 7/29/91)

pkg$images/imarith/imc* -
pkg$images/imarith/ic* +
pkg$images/imarith/t_imcombine.x
pkg$images/imarith/mkpkg
pkg$images/imarith/generic/mkpkg
pkg$images/imcombine.par
pkg$images/doc/imcombine.hlp
    Replaced old version of IMCOMBINE with new version supporting masks,
    offsets, and new algorithms.  (Valdes 7/19/91)

pkg$images/iminfo/imhistogram.x
    Imhistogram has been modified to print the value of the middle of
    histogram bin instead of the left edge if the output type is list 
    instead of plot. (Davis 6/11/91)

pkg$images/t_imsurfit.x
    Modified the sections file reading code to check the order of the
    x1 x2 y1 y2 parameters and switch (x1,x2) or (y1,y2) if x2 < x1 or
    y2 < y1 respectively. (Davis 5/28/91)

pkg$images/listpixels.par
pkg$images/iminfo/listpixels.x
pkg$images/doc/listpixels.hlp
    Modified the listpixels task to be able to print the pixel coordinates
    in logical, physical or world coordinates. The default coordinate
    system is still logical as before. (Davis 5/17/91)

pkg$images/images.par
pkg$images/doc/minmax.hlp
pkg$images/imutil/t_minmax.x
pkg$images/imutil/minmax.x
    Minmax was modified to do the minimum and maximum values computations
    in double precision or complex instead of real if the input image
    pixel type is double precision or complex. Note that the minimum and
    maximum header values are still stored as real however.
    (Davis 5/16/91)

imarith/t_imarith.x
    There was a missing statement to set the error flag if the image
    dimensions did not  match.  (5/14/91, Valdes)

doc/imarith.hlp
    Fixed some formatting problems in the imarith help page. (5/2/91 Davis)

imarith$imcombine.x
    Changed the order in which images are unmapped to have the output images
    closed last.  This is to allow file descriptors for the temporary image
    used when updating STF headers.  (4/22/91, Valdes)

pkg$images/geometry/t_blkavg.x
pkg$images/geometry/blkavg.gx
pkg$images/geometry/blkavg.x
    The blkavg task was partially modified to support complex image data. 
    The full modifications cannot be made because of an error in abavx.x
    and the missing routine absux.x.
    (4/18/91 Davis)

pkg$images/geometry/geofit.x
    The x and y fits cross-terms switch was not being set correctly to "yes"
    in the case where xxorder=2 and xyorder=2 or in the case where yxorder=2
    and yyorder=2.
    (4/9/91 Davis)

pkg$images/geometry/geogmap.x
    Modified the line which prints the geometric parameters to use the
    variable name xshift and yshift instead of delx and dely.
    (4/9/91 Davis)

pkg$images/imfit/imsurfit.x
    Fixed a bug in the pixel rejection code which occurred when upper was >
    0.0 and lower = 0.0 or lower > 0 and upper = 0.0. The problem was that
    the code was simply setting the rejection limits to the computed sigma
    times the upper and lower parameters without checking for the 0.0
    condition first.  In the first case this results in all points with
    negative residuals being rejected and in the latter all points with
    positive residuals are rejected.
    (2/25/91 Davis)

pkg$images/doc/hedit.hlp
pkg$images/doc/hselect.hlp
pkg$images/doc/imheader.hlp
pkg$images/doc/imgets.hlp
    Added a reference to imgets in the SEE ALSO sections of the hedit and
    hselect tasks.
    Added a reference to hselect and hedit in the SEE ALSO sections of the
    imheader and imgets tasks.
    (2/22/91 Davis)

pkg$images/gradient.hlp
pkg$images/laplace.hlp
pkg$images/gauss.hlp
pkg$images/convolve.hlp
pkg$images/gradient.par
pkg$images/laplace.par
pkg$images/gauss.par
pkg$images/convolve.par
pkg$images/t_gradient.x
pkg$images/t_laplace.x
pkg$images/t_gauss.x
pkg$images/t_convolve.x
pkg$images/convolve.x
pkg$images/xyconvolve.x
pkg$images/radcnv.x
    The convolution operators were modified to run more efficiently in
    certain cases. The LAPLACE task was modified to make use of the
    radial symmetry of the convolution kernel in the y direction as well
    as the x direction resulting in a modest speedup in execution time.
    A new parameter bilinear was added to the GAUSS and CONVOLVE tasks.
    By default and if appropriate mathematically,  GAUSS now makes use of
    the bilinearity or separability of the Gaussian function,
    to separate the 2D convolution in x and y into two equivalent
    1D convolutions in x and y, resulting in a considerable speedup
    in execution time. Similarly the user can know program CONVOLVE to
    compute a bilinear convolution instead of a full 2D 1 if appropriate.
    (1/29/91 Davis)

pkg$images/filters/t_convolve.x
    CONVOLVE was not decoding the legal 1D kernel "1.0 2.0 1.0" correctly
    although the alternate form "1.0 2.0 1.0;" worked. Leading
    blanks in string kernels as in for example " 1.0 2.0 1.0" also generated
    and error.  Fixed these bugs and added some  additional error checking code.
    (11/28/90 Davis)

pkg$images/doc/gauss.hlp
    Added a detailed mathematical description of the gaussian kernel used
    by the GAUSS task to the help page.

pkg$images/images.hd
pkg$images/rotate.cl
pkg$images/imlintran.cl
pkg$images/register.cl
pkg$images/register.par
	Added src="script file name" entries to the IMAGES help database
	for the tasks ROTATE, IMLINTRAN, and REGISTER. Changed the CL
	script for REGISTER to a procedure script to remove the ugly
	local variable declarations. Added a few comments to the scripts.
	(12/11/90, Davis)

pkg$images/iminfo/imhistogram.x
    Added a new parameter binwidth to imhistogram. If binwidth is defined
    it determines the histogram resolution in intensity units, otherwise
    nbins determines the resolution as before. (10/26/90, Davis)

pkg$images/doc/sections.hlp
    Clarified what is meant by an image template and that the task itself
    does not check whether the specified names are actually images.
    The examples were improved.  (10/3/90, Valdes)

pkg$images/doc/fit1d.hlp
    Changed lines to columns in example 2.  (10/3/90, Valdes)

pkg$images/imarith/imcscales.x
    When an error occured while parsing the mode section the untrapped error
    caused further problems downstream.  Because it would require adding
    lots of errchks to cause the program to gracefully abort I instead made
    it a warning.  (10/2/90, Valdes)

pkg$images/imutil/hedit.x
    Hedit was computing but not using min_lenarea. If the user specified
    a min_lenuserarea greater than the default of 28800 then the default
    was being used instead of the larger number.

pkg$imarith/imasub.gx
    The case of subtracting an image from the constant zero had a bug
    which is now fixed.  (8/14/90, Valdes)

pkg$images/t_imtrans.x
    Modified the imtranspose task so it will work on type ushort images.
    (6/6/90 Davis)

pkg$images
    Added world coordinate system support to the following tasks: imshift,
    shiftlines, magnify, imtranspose, blkrep, blkavg, rotate, imlintran,
    register and geotran. The only limitation is that register and geotran
    will only support simple linear transformations. 
    (2/24/90 Davis)
    
pkg$images/geometry/geotimtran.x
    Fixed a problem in the boundary extension "reflect" option code for small
    images which was causing odd values to be inserted at the edges of the
    image.
    (2/14/90 Davis)

pkg$images/iminfo/imhistogram.x
    A new parameter "hist_type" was added to the imhistogram task giving
    the user the option of plotting the integral, first derivative and
    second derivative of the histogram as well as the normal histogram.
    Code was contributed by Rob Seaman.
    (2/2/90 Davis)

pkg$images/geometry/geogmap.x
    The path name of the help file was being erroneously renamed with
    the result that when users ran the double precision version of the
    code they could not find the help file.
    (26/1/90 Davis)

pkg$images/filters/t_boxcar.x,t_convolve.x
    Added some checks for 1-D images.
    (1/20/90 Davis)

pkg$images/iminfo/t_imstat.x,imstat.h
    Made several minor bug fixes and alterations in the imstatistics task
    in response to user complaints and suggestions.

    1. Changed the verbose parameter to the format parameter. If format is
    "yes" (the default) then the selected fields are printed in fixed format
    with column labels. Other wise the fields are printed in free format
    separated by 2 blanks. This fixes the problem of fields running together.

    2. Fixed a bug in the code which estimates the median from the image
    histogram by linearly interpolating around the midpt of the integrated
    histogram. The bug occurred when more than half the pixels were in the
    first bin.

    3. Added a check to ensure that the number of fields did not overflow
    the fields array.

    4. Removed the extraneous blank line printed after the title.

    5. The pound sign is now printed at the beginning of the column header
    string regardless of which field is printed first. In the previous
    versions it was only being printed if the image name field was
    printed first.

    6. Changed the name of the median field to midpt in response to user
    confusions about how the median is computed.

    (1/20/90, Davis)

pkg$images/imutil/t_imslice.hlp
    The imslice was not correctly computing the number of lines in the
    output image in the case where the slice dimension was 1.
    (12/4/89, Davis)

pkg$images/doc/imcombine.hlp
    Clarified and documented definitions of the scale, offset, and weights.
    (11/30/89, Valdes)

pkg$images/geometry/geotran.x
    High order surfaces of a certain functional form could occasionally
    produce out of bounds pixel errors. The bug was caused by not properly
    computing the distortion of the image boundary for higher order
    surfaces.
    (11/21/89, Davis)

pkg$images/geometry/imshift.x
    The circulating buffer space was not being freed after each execution
    of IMSHIFT. This did not cause an error in execution but for a long
    list of frames could result in alot of memory being tied up.
    (10/25/89, Davis)

pkg$images/imarith/t_imarith.x
    IMARITH is not prepared to deal with images sections in the output.
    It used to look for '[' to decide if the output specification included
    and image section.  This has been changed to call the IMIO procedure
    imgsection and check if a non-null section string is returned.
    Thus it is up to IMIO to decide what part of the image name is
    an image section.  (9/5/89, Valdes)

pkg$images/imarith/imcmode.gx
    Fixed bug causing infinite loop when computing mode of constant value
    section.  (8/14/89, Valdes)

====
V2.8
====

pkg$images/iminfo/t_imstat.x
    Davis, Jun 15, 1989
    Added a couple of switches to that skew and kurtosis are not computed
    if they are not to be printed.

pkg$images/iminfo/t_imstat.x
    Davis, Jun 14, 1989
    A simple mod was made to the skew and kurtosis computation to avoid
    divide by zero errors in case of underflow.

pkg$images/imutil/chpixtype.par
    Davis, Jun 13, 1989
    The parameter file has been modified to accept an output pixel
    type of ushort.

pkg$images/imarith/imcombine.gx
    Valdes, Jun 2, 1989
    A new scheme to detect file errors is now used.

pkg$images/imfit/t_imsurfit.x
    Davis, Jun 1, 1989
    1. If the user set regions = "sections" but the sections file
    did not exist the task would go into an infinite loop. The problem
    was a missing error check on the open statement.

pkg$images/iminfo/imhistogram.x,imhistogram.par
    Davis, May 31, 1989
    A new version of imhistogram has been installed. These mods have
    been made over a period of a month by Doug Tody and Rob Seaman.
    The mods include
    1. An option to turn off log scaling of the y axis of the histogram plot.
    2. A new autoscale parameter which avoids aliasing problems for integer
       data.
    3. A new parameter top_close which resolves the ambiguity in the top
       bin of the histogram.

pkg$images/imarith/imcombine.gx
   Valdes, May 9, 1989
   Because a file descriptor was not reserved for string buffer operations
   and a call to stropen in cnvdate was not error checked the task would
   hang when more than 115 images were combined.  Better error checking
   was added and now an error message is printed when the maximum number
   of images that can be combined is exceeded.

pkg$images/imarith/t_imarith.x
   Valdes, May 6, 1989
   Operations in which the output image has an image section are now
   skipped with a warning message.

pkg$images/imarith/sigma.gx
pkg$images/imarith/imcmode.gx
    Valdes, May 6, 1989
    1.  The weighted sigma was being computed incorrectly.
    2.  The argument declarations were wrong for integer input images.
	Namely the mean vector is always real.
    3.  Minor change to imcmode.gx to return correct datatype.

pkg$images/imstack,imslice
    Davis, April 1, 1989
    The proto images tasks imstack and imslice have been moved from the
    proto package to the images package. Imstack is unchanged except that
    it now supports the image data types USHORT and COMPLEX. Imslice has
    been modified to allow slicing along any dimension of the image instead
    of just the highest dimension.

pkg$images/imstatistics.
    Davis, Mar 31, 1989
    1. A totally new version of the imstatistics task has been written
    and replaces the old version. The new task allows the user to select
    which statistical parameters to compute and print. These include
    the mean, median, mode, standard deviation, skew, kurtosis and the
    minimum and maximum pixel values.

pkg$images/imhistogram.par
pkg$images/iminfo/imhistogram.x
pkg$images/doc/imhistogram.hlp
    Davis, Mar 31, 1989
    1. The imhistogram task has been modified to plot "box" style histograms
    as well as "line" type histograms. Type "line" remains the default.

pkg$images/geometry/geotran.par,register.par,geomap.par
pkg$images/doc/geomap.hlp,register.hlp,geotran.hlp
    Davis, Mar 6, 1989
    1. Improved the parameter prompting in GEOMAP, REGISTER and GEOTRAN
    and improved the help pages.
    2. Changed GEOMAP database quantities "xscale" and "yscale" to "xmag"
    and "ymag" for consistency . Geotran was changed appropriately.

pkg$images/imarith/imcmode.gx
    For short data a short variable was wraping around when there were
    a significant number of saturated pixels leading to an infinite loop.
    The variables were made real regardless of the image datatype.
    (3/1/89, Valdes)
 
pkg$images/imutil/imcopy.x
    Davis, Feb 28, 1989
    1. Added support for type USHORT to the imcopy task. This is a merged
    ST modification.

pkg$images/imarith/imcthreshold.gx
pkg$images/imcombine.par
pkg$images/doc/imcombine.hlp
pkg$images/imarith/imcscales.x
    Valdes, Feb 16, 1989
    1.  Added provision for blank value when all pixels are rejected by the
        threshold.
    2.  Fixed a bug that was improperly scaling images in the threshold option.
    3.  The offset printed in the log now has the opposite sign so that it
	is the value "added" to bring images to a common level.

pkg$images/imfit/imsurfit.x
    Davis, Feb 23, 1989
    Fixed a bug in the median fitting code which could cause the porgram
    to go into an infinite loop if the region to be fitted was less than
    the size of the whole image.

pkg$images/geometry/t_magnify.x
    Davis, Feb 16, 1989
    Modified magnify to work on 1D images as well as 2D images. The
    documentation has been updated.

pkg$images/geometry/t_geotran.x
    Davis, Feb 15, 1989
    Modified the GEOTRAN and REGISTER tasks so that they can handle a list
    of transform records one for each input image.

pkg$images/imarith/imcmode.gx
    Valdes, Feb 8, 1989
    Added test for nx=1.
 
pkg$images/imarith/t_imcombine.x
    Valdes, Feb 3, 1989
    The test for the datatype of the output sigma image was wrong.
 
pkg$images/iminfo/listpixels.x,listpixels.par
    Davis, Feb 6, 1989
    The listpixels task has been modified to print out the pixels for a 
    list of images instead of a single image only. A title line for each
    image listed can optionally be printed on the standard output if
    the new parameter verbose is set to yes.

pkg$images/geometry/t_imshift.x
    Davis, Feb 2, 1989
    Added a new parameter shifts_file to the imshift task. Shifts_file
    is the name of a text file containing the the x and yshifts for
    each input image to be shifted. The number of input shifts must
    equal the number of input images.

pkg$images/geometry/t_geomap.x
    Davis, Jan 17, 1989
    Added an error message for the case where the coordinates is empty
    of there are no points in the specified data range. Previously the
    task would proceed to the next coordinate file without any message.

pkg$images/geometry/t_magnify.x
    Davis, Jan 14, 1989
    Added the parameter flux conserve to the magnify task to bring it into
    line with all the other geometric transformation tasks.

pgk$images/geometry/geotran.x,geotimtran.x
    Davis, Jan 2, 1989
    A bug was fixed in the flux conserve code. If the x and y reference
    coordinates  are not in pixel units and are not 1 then
    the computed flux per pixel was too small by xscale * yscale.

pkg$images/filters/acnvrr.x,convolve.x,boxcar.x,aboxcar.x
    Davis, Dec 27, 1988
    I changed the name of the acnvrr procedure to cnv_radcnvr to avoid
    a name conflict with a vops library procedure. This only showed
    up when shared libraries were implemented. I also changed the name
    of the aboxcarr procedure to cnv_aboxr to avoid conflict with the
    vops naming conventions.

pkg$images/imarith/imcaverage.gx
    Davis, Dec 22, 1988
    Added an errchk statement for imc_scales and imgnl$t to stop the
    program bombing with segmentation violations when mode <= 0.

pkg$images/imarith/imcscales.x
    Valdes, Dec 8, 1988
    1.  IMCOMBINE now prints the scale as a multiplicative quantity.
    2.  The combined exposure time was not being scaled by the scaling
	factors resulting in a final exposure time inconsistent with the
	data.

pkg$images/iminfo/imhistogram.x
    Davis, Nov 30, 1988
    Changed the list+ mode so that bin value and count are printed out instead
    of bin count and value. This makes the plot and list modes compatable.

pkg$images/iminfo/t_imstat.x
    Davis, Nov 17, 1988
    Added the n=n+1 back into the inner loop of imstat.

pkg$images/geotran.par,register.par
    Davis, Nov 11 , 1988
    Fixed to glaring errors in the parameter files for register and geotran.
    Xscale and yscale were described as pixels per reference unit when
    they should be reference units per pixel. The appropriate bug fix has been
    made.

pkg$images/geometry/t_geotran.x
    Davis, November 7, 1988
    The routine gsrestore was not being error checked. If either of the
    input x or y coordinate surface was linear and the other was not,
    the message came back GSRESTORE: Illegal x coordinate. This bug has
    been fixed.

pkg$images/imarith/imcombine.gx
    Valdes, October 19, 1988
    A vops clear routine was not called generically causing a crash with
    double images.

pkg$images/filters/t_fmedian.x,t_median.x,t_fmode.x,t_mode.x,t_gradient.x
    t_gauss.x,t_boxcar.x,t_convolve.x,t_laplace.x
    Davis, October 4, 1988
    I fixed a bug in the error handling code for the filters tasks. If
    and error occurred during task execution and the input image name was
    the same as the output image name then the input image was trashed.

pkg$images/imarith/imcscales.gx
    Valdes, September 28, 1988
    It is now an error for the mode to be nonpositive when scaling or weighting.

pkg$images/imarith/imcmedian.gx
    Valdes, August 16, 1988
    The median option was selecting the n/2 value instead of (n+1)/2.  Thus,
    for an odd number of images the wrong value was being determined for the
    median.

pkg$images/geometry/t_imshift.x
    Davis, August 11, 1988
    1. Imshift has been modified to uses the optimized code if nearest
    neighbour interpolation is requested. A nint is done on the shifts
    before calling the quick shift routine.
    2. If the requested pixel shift is too large imshift will now
    clean up any pixelless header files before continuing execution.

pkg$images/geometry/blkavg.gx
    Davis, July 13, 1988
    Blkavg has been fixed so that it will work on 1D images.

pkg$images/geometry/t_imtrans.x,imtrans.x
    Davis, July 12, 1988
    Imtranspose has been modified to work on complex images.

pkg$images/imutil/t_chpix.x
    Davis, June 29, 1988
    A new task chpixtype has been added to the images package. Chpixtype
    changes the pixel types of a list of images to a specified output pixel
    type. Seven data types are supported "short", "ushort", "int", "long"
    "real" and "double".

pkg$images/geometry/rotate.cl,imlintran.cl,t_geotran.x
    Davis, June 10, 1988
    The rotate and imlintran scripts have been rewritten to use procedure
    scripts. This removes all the annoying temporary cl variables which
    appear when the user does an lpar. In previous versions of these
    two tasks the output was restricted to being the same size as the input
    image. This is still the default case, but the user can now set the
    ncols and nrows parameters to the desired output size. I ncols or nlines
    < 0 then then the task compute the output image size required to contain
    the whole input image.

pkg$images/filters/t_convolve.x,t_laplace.x,t_gradient.x,t_gauss.x,convolve.x
    Davis, June 1, 1988
    The convolution operators laplace, gauss and convolve have been modified
    to make use of radial symmetry in the convolution kernel. In gauss and
    laplace the change is transparent to the user. For the convolve operator
    the user must indicate that the kernel is radially symmetric by setting
    the parameter radsym. For kernels of 7 by 7 or greater the speedup
    in timings is on the order of 30% on the Vax 750 with the fpa.

pkg$images/imarith/imcmode.gx
    Valdes, Apr 11, 1988
    1.  The use of a mode sections was handled incorrectly.

pkg$images/imfit/fit1d.x
    Valdes, Jan 4, 1988
    1.  Added an error check for a failure in IMMAP.  The missing error check
	caused FIT1D to hang when a bad input image was specified.

pkg$images/magnify.par
pkg$images/imcombine.par
pkg$images/imarith/imcmode.gx
pkg$images/doc/imarith.hlp
    Valdes, Dec 7, 1987
    1.  Added option list to parameter prompts.
    2.  Fixed minor typo in help page
    3.  The mode calculation in IMCOMBINE would go into an infinite loop
	if all the pixel values were the same.  If all the pixels are the
	same them it skips searching for the mode and returns the constant
	number.

pkg$images/geometry/geotimtran.x
    Davis, Nov 25, 1987
    1. A bug in the boundary extension = wrap option was found in the
    IMLINTRAN task. The problem occured in computing values for out of
    bounds pixels in the range 0.0 < x < 1.0, ncols < x < ncols + 1.0,
    0.0 < y < 1.0 and nlines < y < nlines + 1.  The computed coordinates
    were falling outside the boundaries of the interpolation array.

pkg$images/geometry/t_geomap.x,geograph.x
    Davis, Nov 19, 1987
    1. The geomap task now writes the name of the output file into the database.
    2. Rotation angles of 360. degrees have been altered to 0 degrees.

pkg$images/imfit/t_imsurfit.x,imsurfit.x
pkg$images/lib/ranges.x
    Davis, Nov 2, 1987
    A bug in the regions fitting option of the IMSURFIT task has been found
    and fixed. This bug would occur when the user set the regions parameter
    to sections and then listed section which overlapped each other. The
    modified ranges package was not handling the overlap correctly and
    computing a number of points which was incorrect.

pkg$images/imarith/* +
    Valdes, Sep 30, 1987
    The directory was reorganized to put generic code in the subdirectory
    generic.

    A new task called IMCOMBINE has been added.  It provides for combining
    images by a number of algorithms, statistically weighting the images
    when averaging, scaling or offsetting the images by the exposure time
    or image mode before combining, and rejecting deviant pixels.  It is
    almost fully generic including complex images and works on images of
    any dimension.

pkg$images/geometry/geotran.x
    Davis, Sept 3, 1987
    A bug in the flux conserving algorithm was found in the geotran code.
    The symptom was that the flux of the output image occasionally was
    negative. This would happen when two conditions were met, the transformation
    was of higher order than a simple rotation, magnification, translation
    and an axis flip was involved. The mathematical interpretation of this
    bug is that the coordinate surface had turned upside down. The solution
    for people running systems with this bug is to multiply there images
    by -1.

pkg$images/imfit/imsurfit.h,t_imsurfit.x
    Davis, Aug 6, 1987
    A new option was added to the parameter regions in the imsurfit task.
    Imsurfit will now fit a surface to a single circular region defined
    by an x and y center and a radius.

pkg$images/geometry/geotimtran.x
    Davis, Jun 15, 1987
    Geotran and register were failing when the output image number of rows
    and columns was different from the input number of rows and columns.
    Geotran was mistakenly using the input images sizes to determine the
    number of output lines that should be produced. The same problem occurred
    when the values of the boundary pixels were being computed. The program
    was using the output image dimensions to compute the boundary pixels
    instead of the input image dimensions.

pkg$images/geometry/geofit.x,geogmap.x
    Davis, Jun 11, 1987
    A bug in the error checking code in the geomap task was fixed. The
    condition of too few points for a reasonable was not being trapped
    correctly. The appropriate errchk statements were added.

pkg$images/geomap.par
    Davis, Jun 10, 1987
    The default fitting function was changed to polynomial. This will satisfy
    most users who wish to do shifts, rotations, and magnifications and
    avoid the neccessity of correctly setting the xmin, xmax, ymin, and ymax
    parameters. For the chebyshev and legendre polynomial functions these
    parameters must be explicitly set.  For reference coordinates in pixel
    units the normal settings are 1, ncols, 1 and nlines respectively.

pkg$images/iminfo/hselect.x,imheader.x,images$/imutil/hselect.x
    Davis, Jun 8, 1987
    Imheader has been modified to open an image with the default min_lenuserarea
    Hselect and hedit will now open the image setting the user area to the
    maximum of 28800 chars or the min_lenuser environment variable.

pkg$images/iminfo/t_imstat.x
    Davis, May 22, 1987
    An error in the image minimum computation was corrected. This error
    would show up most noiticeably if imstat was run on a 1 pixel image.
    The min value would be left set to MAX_REAL.

pkg$images/filters/mkpkg
    Davis, May 22, 1987
    I added mach.h to the dependency file list of t_fmedian.x and
    recompiled. The segmentation violations I had been getting in the
    program disappeared.

pkg$images/t_shiftlines.x,shiftlines.x
    Davis, April 15, 1987
    1. I changed the names of the procedures shiftlines and shiftlinesi
    to sh_lines and sh_linesi. When the original names were contracted
    to 6 letter fortran names they became shifti and shifts which just
    so happens to collide with shifti and shifts in the subdirectory
    osb. On VMS this was causing problems with the shareable libraries.
    If images was linked with -z there was no problem.

pkg$images/imarith/t_imsum.x
    Valdes, March 24, 1987
    1.  IMSUM was failing to unmap images opened to check image dimensions
	in a quick first pass through the image list.  This is probably
	the source of the out of files problem with STF images.  It may
	be the source of the out of memory problem reported from AOS/IRAF.

pkg$images/imfit/fit1d.x
pkg$images/imfit/mkpkg
    Valdes, March 17, 1987
    1. Added error checking for the illegal operation in which both input
       and output image had an image section.  This was causing the task
       to crash.  The task now behaves properly in this circumstance and
       even allows the fitted output to be placed in an image section of
       an existing output image (even different than the input image
       section) provided the input and output images have the same sizes.

pkg$images/t_convolve.x
    Davis, March 3, 1987
    1. Fixed the kernel decoding routine in the convolve task so that
    it now recognizes the row delimter character in string entry mode.

pkg$images/geometry,filters
    Davis, February 27, 1987
    1. Changed all the imseti (im, TY_BNDRYPIXVAL, value) calls to imsetr.

pkg$images/t_minmax.x,minmax.x
    Davis, February 24, 1987
    1. Minmax has been changed to compute the minimum and maximum pixel
    as well as the minimum and maximum pixel values. The pixels are output
    in section notation and stored in the minmax parameter file.

pkg$images/t_magnify.x
    Davis, February 19, 1987
    1. Magnify was aborting with the error MSIFIT: Too few datapoints
    when trying to reduce an image using the higher order interpolants
    poly3, poly5 and spline3. I increased the NEDGE defined constant
    from 2 to three and modified the code to use the out of bounds
    imio.

pkg$images/geograph.x,geogmap.x
    Davis, February 17, 1987
    1. Geomap now uses the gpagefile routine to page the .keys file.
    The :show command deactivates the workstation before printing a
    block of text and reactivates it when it is finished.

pkg$images/geometry/geomap,geotran
    Davis, January 26, 1987
    1. There have been substantial changes to the geomap, and geotrans
    tasks and those tasks rotate, imlintran and register which depend
    on them.
    2. Geomap has been changed to be able to compute a transformation
    in both single and double precision.
    3. The geotran code has been speeded up considerably. A simple rotate
    now takes 70 seconds instead of 155 seconds using bilinear interpolation.
    4. Two new cl parameters nxblock and nyblock have been added to the
    rotate, imlintran, register and geotran tasks. If the output image
    is smaller than these parameters then the entire output image
    is computed at once. Otherwise the output image is computed in blocks
    nxblock by nyblock in size.
    5. The 3 geotran parameters rotation, scangle and flip have been replaced
    with two parameters xrotation and yrotation which serve the same purpose.

pkg$images/geometry/t_shiftlines.x,shiftlines.x
    Davis, January 19, 1987
    1. The shiftlines task has been completely rewritten. The following
    are the major changes.
    2. Shiftlines now makes use of the imio boundary extension operations.
    Therefore the four options: nearest pixel, reflect, wrap and constant
    boundary extension are available.
    3. The interpolation code has been vectorised. The previous version
    was using the function call asieval for every output pixel evaluated.
    The asieval call were replaced with asivector calls.
    4. An extra CL parameter constant to support constant boundary
    exension was added.
    5. The shiftlines help page was modified and the date changed to
    January 1987.

pkg$images/imfit/imsurfit.x
    Davis, January 12, 1987
    1. I changed the amedr call to asokr calls. For my application it did
    not matter whether the input array is left partially sorted and the asokr
    routine is more efficient.

pkg$images/lib/pixlist.x
    Davis, December 12, 1986
    1. A bug in the pl_get_ranges routine caused the routine to fail when the
    number of ranges got too large. The program could not detect the end of
    the ranges and would go into an infinite loop.

pkg$images/iminfo/t_imstat.x
    Davis, December 3, 1986
    1. Imstat was failing on constant images because finite machine precision
    could result in a negative sigma squared. Added a check for this condition.

pkg$images/filters/fmode.x
    Davis, October 27, 1986
    1. Added a check for 0 data range before calling amapr.

pkg$images/imarith/imsum.gx
    Valdes, October 20, 1986
    1. Found and fixed bug in this routine which caused pixel rejection
    to fail some fraction of the time.

pkg$images/geometry/blkrp.gx
    Valdes, October 13, 1986
    1.  There was a bug when the replication factor for axis 1 was 1.

pkg$images/iminfo/imhistogram.x
    Hammond, October 8, 1986
    1. Running imhistogram on a constant valued image would result in
       a "floating divide by zero fault" in ahgm.  This condition is
       now trapped and a warning printed if there is no range in the data.

pkg$images/tv/doc/cvl.hlp
    Valdes, October 7, 1986
    1.  Typo in V2.3 documentation fixed: "zcale" -> "zscale".

pkg$images/fit1d.par
    Valdes, October 7, 1986
    1.  When querying for the output type the query was:

Type of output (fit, difference, ratio) (fit|difference|ratio) ():

    	The enumerated values were removed since they are given in the
	prompt string.

pkg$images/imarith/t_imsum.x
pkg$images/imarith/imsum.gx
pkg$images/do/imsum.hlp
    Valdes, October 7, 1986
    1.  Medians or pixel rejection with more than 15 images is now
	correct.  There was an error in buffering.
    2.  Averages of integer datatype images are now correct.  The error
	was caused by summing the pixel values divided by the number
	of images instead of summing the pixel values and then dividing
	by the number of images.
    3.  Option keywords may now be abbreviated.
    4.  The output pixel datatype now defaults to the calculation datatype
	as is done in IMARITH.  The help page was modified to indicate this.
    5.  Dynamic memory is now used throughout to reduce the size of the
	executable.
    6.  The bugs 1-2 are present in V2.3 and not in V2.2.

pkg$images/imarith/t_imarith.x
pkg$images/imarith.par
pkg$images/doc/imarith.hlp
    Valdes, October 6, 1986
    1.  The parameter "debug" was changed to "noact".  "debug" is reserved
	for debugging information.
    2.  The output pixel type now defaults to the calculation datatype.
    3.  The datatype of constant operands is determined with LEXNUM.  This
	fixes a bug in which a constant such as "1." was classified as an
	integer.
    4.  Trailing whitespace in the string for a constant operand is allowed.
	This fixes a bug with using "@" files created with the task FIELDS
	from a table of numbers.  Trailing whitespace in image names is
	not checked for since this should be taken care of by lower level
	system services.
    5.  The reported bug with the "max" operation not creating a pixel file
	was the result of the previous round of changes.  This has been
	corrected.  This problem does not exist in the released version.
    6.  All strings are now dynamically allocated.  Also IMTOPENP is used
	to open a CL list directly.
    7.  The help page was revised for points (1) and (2).

pkg$images/fmode.par
pkg$images/fmd_buf.x
pkg$images/med_sort.x
    Davis, September 29, 1986
    1. Changed the default value of the unmap parameter in fmode to yes. The
       documentation was changed and the date modified.
    2. Added a test to make sure that the input image was not a constant
       image in fmode and fmedian.
    3. Fixed the recently added swap macro in the sort routines which
       was giving erroneous results for small boxes in tasks median and mode.

pkg$images/imfit/fit1d.x
    Valdes, September 24, 1986
    1.  Changed subroutine name with a VOPS prefix to one with a FIT1D
	prefix.

pkg$images/imarith/t_imdivide.x
pkg$images/doc/imdivide.hlp
pkg$images/imdivide.par
    Valdes, September 24, 1986
    1.  Modified this ancient and obsolete task to remove redundant
	subroutines now available in the VOPS library.
    2.  The option to select action on zero divide was removed since
	there was only one option.  Parameter file changed.
    3.  Help page revised.

pkg$images/geometry/t_blkrep.x +
pkg$images/geometry/blkrp.gx +
pkg$images/geometry/blkrep.x +
pkg$images/doc/blkrep.hlp +
pkg$images/doc/mkpkg
pkg$images/images.cl
pkg$images/images.men
pkg$images/images.hd
pkg$images/x_images.x
    Valdes, September 24, 1986
    1.  A new task called BLKREP for block replicating images has been added.
	This task is a complement to BLKAVG and performs a function not
	available in any other way.
    2.  Help for BLKREP has been added.

pkg$images/imarith/t_imarith.x
pkg$images/imarith/imadiv.gx
pkg$images/doc/imarith.hlp
pkg$images/imarith.par
    Valdes, September 24, 1986
    1.  IMARITH has been modified to provide replacement of divisions
	by zero with a constant parameter value.
    2.  The documentation has been revised to include this change and to
	clarify and emphasize areas of possible confusion.

pkg$images/doc/magnify.hlp
pkg$images/doc/blkavg.hlp
    Valdes, September 18, 1986
    1.  The MAGNIFY help document was expanded to clarify that images with axis
        lengths of 1 cannot be magnified.  Also a discussion of the output
	size of a magnified image.  This has been misunderstood often.
    2.  Minor typo fix for BLKAVG.

images$geometry/blkav.gx: Davis, September 7, 1986
    1. The routine blkav$t was declared a function but called everywhere as
    a procedure. Removed the function declaration.

images$filters/med_sort.x: Davis, August 14, 1986
    1. A bug in the sorting routine for MEDIAN and MODE in which the doop
    loop increment was being set to zero has been fixed. This bug was
    causing MEDIAN and MODE to fail on class 6 for certain sized windows.

images$imfit/fit1d.x: Davis, July 24, 1986
    1. A bug in the type=ratio option of fit1d was fixed. The iferr call
    on the vector operator adivr was not trapping a divide by zero
    condition. Changed adivr to adivzr.

images$iminfo/listpixels.x: Davis, July 21, 1986
    1. I changed a pargl to pargi for writing out the column number of the
    pixels.

images$iminfo/t_imstat.x: Davis, July 21, 1986
    1. I changed a pargr to a pargd for the double precision quantitiies
    sum(MIN) and sum(MAX).

images$imfit/t_lineclean.x: Davis, July 14, 1986
    1. Bug in the calling sequence for ic_clean fixed. The ic pointer
    was not being passed to ic_clean causing access violation and/or
    segmentation violation errors.

images$imfit/fit1d.x, lineclean.x:  Valdes, July 3, 1986
    1.  FIT1D and LINECLEAN modified to use new ICFIT package.

From Valdes June 19, 1986

1. The help page for IMSUM was modified to explicitly state what the
median of an even number of images does.

-----------------------------------------------------------------------------

From Davis June 13, 1986

1. A bug in CONVOLVE in which insufficient space was being allocated for
long (> 161 elements) 1D kernels has been fixed. CONVOLVE was not
allocating sufficent extra space.

-----------------------------------------------------------------------------

From Davis June 12, 1986

1. I have changed the default value of parameter unmap in task FMEDIAN to
yes to preserve the original data range.

2. I have changed the value of parameter row_delimiter from \n to ;.

-----------------------------------------------------------------------------

From Davis May 12, 1986

1. Changed the angle convention in GAUSS so that theta is the angle of the
major axis with respect to the x axis measured counter-clockwise as specified
in the help page instead of the negative of that angle.

-----------------------------------------------------------------------------

From Davis Apr 28, 1986

1. Moved geomap.key to lib$scr and made redefined HELPFILE in geogmap.x
appropriately.

------------------------------------------------------------------------------

images$imarith/imsum.gx:  Valdes Apr 25, 1986
    1.  Fixed bug in generic code which called the real VOPS operator
	regardless of the datatype.  This caused IMSUM to fail on short
	images.

From Davis Apr 17, 1986

1. Changed constructs of the form boolean == false in the file imdelete.x
to ! boolean.

------------------------------------------------------------------------------

images$imarith:  Valdes, April 8, 1986
    1.  IMARITH has been modified to also operate on a list of specified
	header parameters.  This is primarily used when adding images to
	also added the exposure times.  A new parameter was added and the
	help page modified.
    2.  IMSUM has been modified to also operate on a list of specified
	header parameters.  This is primarily used when summing images to
	also sum the exposure times.  A new parameter was added and the
	help page modified.

------------------------------------------------------------------------------

From Valdes Mar 24, 1986:

1.  When modifying IMARITH to handle mixed dimensions the output image header
was made a copy of the image with the higher dimension.  However, the default
when the images were of the same dimension changed to be a copy of the second
operand.  This has been changed back to being a copy of the first operand
image.

------------------------------------------------------------------------------

From Davis Mar 21, 1986:

1. A NULL pointer bug in the subroutine plfree inside IMSURFIT was causing
segmentation violation errors. A null pointer test was added to plfree.

------------------------------------------------------------------------------

From Davis Mar 20, 1986:

1. A bug involving in place operations in several image tasks has been  fixed.

------------------------------------------------------------------------------

From Davis Mar 19, 1986:

1. IMSURFIT no longer permits the input image to be replaced by the output
image.

2. The tasks IMSHIFT, IMTRANSPOSE, SHIFTLINES, and GEOTRAN  have been modified
to use the images tools xt_mkimtemp and xt_delimtemp for in place
calculations.

-------------------------------------------------------------------------------

From Valdes Mar 13, 1986:

1.  Bug dealing with type coercion in short datatype images in IMARITH and IMSUM
which occurs on the SUN has been fixed.
------
From Valdes Mar 10, 1986:

1.  IMSUM has been modified to work on any number of images.

2.  Modified the help page
------
From Valdes Feb 25, 1986:

There have been two changes to IMARITH:

1.  A bug preventing use of image sections has been removed.

2.  An improvement allowing use of images of different dimension.
The algorithm is as follow:

	a.  Check if both operands are images.  If not the output
	image is a copy of the operand image.

	b.  Check that the axes lengths are the same for the dimensions
	in common.  For example a 3D and 2D image must have the same
	number of columns and lines.

	c.  Set the output image to be a copy of the image with the
	higher dimension.

	d.  Repeat the operation over the lower dimensions for each of
	the higher dimensions.

For example, consider subtracting a 2D image from a 3D image.  The output
image will be 3D and the 2D image is subtracted from each band of the
3D image.  This will work for any combination of dimensions.  Another
example is dividing a 3D image by a 1D image.  Then each line of each
plane and each band will be divided by the 1D image.  Likely applications
will be subtracting biases and darks and dividing by response calibrations
in stacked observations.

3.  Modified the help page
===========
Release 2.2
===========
From Davis Mar 6, 1986:

1. A serious bug had crept into GAUSS after I made some changes. For 2D
images the sense of the sigma was reversed, i.e sigma = 2.0 was actually
sigma = 0.5. This bug has now been fixed.

---------------------------------------------------------------------------

From Davis Jan 13, 1986:

1. Listpixels will now print out complex pixel values correctly.

---------------------------------------------------------------------------

From Davis Dec 12, 1985:

1. The directional gradient operator has been added to the images package.

---------------------------------------------------------------------------

From Valdes Dec 11, 1985:

1.  IMARITH has been modified to first check if an operand is an existing
file.  This allows purely numeric image names to be used.

---------------------------------------------------------------------------

From Davis Dec 11, 1985:

1. A Laplacian (second derivatives) operator has been added to the images
package.

---------------------------------------------------------------------------

From Davis Dec 10, 1985:

1. The new convolution tasks  boxcar, gauss and convolve have been added
to the images package. Convolve convolves an image with an arbitrary
user supplied rectangular kernel. Gauss convolves an image with a 2D
Gaussian of arbitrary size. Boxcar will smooth an image using a smoothing
window of arbitrary size.

2. The images package source code has been reorganized into the following
subdirectories: 1) filters 2) geometry 3) imfit 4) imarith 4) iminfo and
5) imutil 6) lib. Lib contains routines which may be of use to several IRAF
tasks such as ranges. The imutil subdirectory contains tasks which modify
images in some way such as hedit. The iminfo subdirectory contains code
for displaying header and pixel values and other image characteristics
such as the histogram. Image arithmetic and fitting routines are found
in imarith and imfit respectively. Filters contains the convolution and
median filtering routines and geometry contains the geometric distortion
corrections routines.

3. The documentation of the main images package has been brought into
conformity with the new IRAF standards.

4. Documentation for imdelete, imheader, imhistogram, listpixels and
sections has been added to the help database.

5. The parameter structure for imhistogram has been simplified. The
redundant parameters sections and setranges have been removed.

---------------------------------------------------------------------------


From Valdes Nov 4, 1985:

1.  IMCOPY modified so that the output image may be a directory.  Previously
logical directories were not correctly identified.
------

From Davis Oct 21, 1985:

1. A bug in the pixel rejection cycle of IMSURFIT was corrected. The routine
make_ranges in ranges.x was not successfully converting a sorted list of
rejected pixels into a list of ranges in all cases.

2. Automatic zero divide error checking has been added to IMSURFIT.
------
From Valdes Oct 17, 1985:

1.  Fit1d now allows averaging of image lines or columns when interactively
setting the fitting parameters.  The syntax is "Fit line = 10 30"; i.e.
blank separated line or column numbers.  A single number selects just one
line or column.  Be aware however, that the actual fitting of the image
is still done on each column or line individually.

2.  The zero line in the interactive curve fitting graphs has been removed.
This zero line interfered with fitting data near zero.
------
From Rooke Oct 10, 1985:

1.  Blkaverage was changed to "blkavg" and modified to support any allowed
number of dimensions.  It was also made faster in most cases, depending on 
the blocking factors in each dimension.
------
From Valdes Oct 4, 1985:

1.  Fit1d and lineclean modified to allow separate low and high rejection
limits and rejection iterations.
------
From Davis Oct 3, 1985:

1. Minmax was not calculating the minimum correctly for integer images.
because the initial values were not being set correctly.
------
From Valdes Oct 1, 1985:

1. Imheader was modified to print the image history.  Though the history
mechanism is little used at the moment it should become an important part
of any image.

2.  Task revisions renamed to revs.
------
From Davis Sept 30, 1985:

1. Two new tasks median and fmedian have been added to the images package.
Fmedian is a fast median filtering algorithm for integer data which uses
the histogram of the image to calculate the median at each window. Median
is a slower but more general algorithm which performs the same task.
------
From Valdes August 26, 1985:

1.  Blkaverage has been modified to include an new parameter called option.
The current options are to average the blocks or sum the blocks.
------
From Valdes August 7, 1985

1.  Fit1d and lineclean wer recompiled with the modified icfit package.
The new package contains better labeling and graph documentation.

2.  The two tasks now have parameters for setting the graphics device
and reading cursor input from a file.
______
From: /u2/davis/ Tue 08:27:09 06-Aug-85
Package: images
Title: imshift bug
 
Imshift was shifting incorrectly when an integral pixel shift in x and
a fractional pixel shift in y was requested. The actual x shift was
xshift + 1. The bug has been fixed and imshift will now work correctly for
any combination of fractional and integral pixel shifts
------
From: /u2/davis/ Fri 18:14:12 02-Aug-85
Package: images
Title: new images task
 
A new task GEOMAP has been added to the images package. GEOMAP calculates
the spatial transformation required to map one image onto another.
------
From: /u2/davis/ Thu 16:47:49 01-Aug-85
Package: images
Title: new images tasks
 
The tasks ROTATE, IMLINTRAN and GEODISTRAN have been added to the images
package. ROTATE rotates and shifts an image. IMLINTRAN will rotate, rescale
and shift an an image. GEODISTRAN corrects an image for geometric distortion.
------
From Valdes July 26, 1985:

1.  The task revisions has been added to page revisions to the images
package.  The intent is that each package will have a revisions task.
Note that this means there may be multiple tasks named revisions loaded
at one time.  Typing revisions alone will give the revisions for the
current package.  To get the system revisions type system.revisions.

2.  A new task called fit1d replaces linefit.  It is essentially the same
as linefit except for an extra parameter "axis" which selects the axis along
which the functions are to be fit.  Axis 1 is lines and axis 2 is columns.
The advantages of this change are:

	a.  Column fitting can now be done without transposing the image.
	    This allows linefit to be used with image sections along
	    both axes.
	b.  For 1D images there is no prompt for the line number.
.endhelp
