Manpage of immerge
Section: User Commands (1)
Return to Main Contents
immerge - Linear merging of images.
IMMERGE is a Miriad task to linearly merge together two images with
different resolutions. The two images must be of the same field
and use the same coordinate system.
In combining the data, it is assumed that the low resolution
image better represents the short spacing data, whereas the
high resolution best represents the fine structure. Commonly,
the low resolution image will be a single-dish observation,
and the high resolution will be a mosaiced interferometric
This gives the two images to be merged. The first input must be
the high resolution image, and the second the low. There is
no default. The two images must be on the same coordinate grid.
If necessary, use REGRID to achieve this.
The output image. No default.
The flux calibration factor. Ideally the two inputs will have
correctly calibrated flux scales, in units like Jy/beam. In this
case, the flux calibration factor would be 1. In practise, the
calibration may not be perfect, and an extra calibration factor
will be required to align the flux scales. "factor" gives the
factor to scale the low resolution image to put it on the same
flux scale as the high resolution image. Note that this factor is
in addition to the scaling needed to convert Jy/beam at a low
resolution to Jy/beam at the high resolution.
If no value is given for "factor", IMMERGE determines this by
comparing the values in the Fourier plane, of data within an
particular annulus (after accounting for the differing resolutions
of the two inputs). IMMERGE finds the scale factor which
minimises differences between the data of the two (in a robust/L1
sense) in the annulus.
This specifies an annulus, in the Fourier domain, where the
high and low resolution images should agree (after allowing
for resolution differences). This is the annulus of data used to
deduce the flux calibration factor, and with options=feather.
Two or three values can be given. The first two give the inner
and outer radius of the annulus in the Fourier domain. The third
argument gives the units. Possible units are "klambda"
(kilo-wavelengths), "meters", "feet" and "nanoseconds". The default
Values for "uvrange" must be given either if "options=feather"
is used or if the flux calibration factor is being deduced.
Region-of-interest parameter. See the help on ``region''
for more information. NOTE: This parameter is ONLY used for
determining the flux calibration factor. Only plane selection
(e.g. via the ``image'' command) is allowed. Typically you would
want to select a range of planes which contains significant signal
in the overlap region.
PGPLOT device for a plot. When determining the flux calibration
factor, IMMERGE can produce a plot showing the correspondence
between the high and low resolution data points in the annulus
(after correcting for resolution effects and the deduced flux
calibration factor). Ideally it will show a line with "y=x".
The default is not to produce a plot. It also plots the
difference from this "y=x" line as a function of spatial frequency.
Before Fourier transforming, the images are padded with a guard
band. "guard" gives one or two values, being the minimum width of
this guard band in pixels, in the x and y directions. The actual
guard band used is such that the size of the image plus guard band
is a power of 2.
Task enrichment parameters. Several can be given, separated by commas.
Minimum match is used.
normalize Rather than the output being the merged images, the
output is the low resolution image corrected by
the flux calibration factor.
zero By default, IMMERGE pads the guard band with data which
minimizes FFT edge effects. If the input images are
really zero beyond the edges of the two input images,
then padding with zeros might be preferable. This is
particularly so if IMMERGE is deducing the flux
calibration scale factor.
feather This merges the two images together in a fashion similar
to AIPS IMERG. This method is generally less desirable
than the default scheme used by IMMERGE.
shift Determine the optimum shift to apply to the low
resolution image to make it align with the high
notaper Normally the low-resolution image is tapered to match any
residual primary beam response in the high-resolution image.
This option causes this step to be skipped.
- PERSON RESPONSIBLE
This document was created by
using the manual pages.
Time: 18:35:38 GMT, July 05, 2011