Документ взят из кэша поисковой машины. Адрес оригинального документа : http://xmm.vilspa.esa.es/sas/5.4.1/doc/asmooth/node6.html
Дата изменения: Fri Jan 10 22:48:37 2003
Дата индексирования: Sat Dec 22 08:11:30 2007
Кодировка:

Поисковые слова: cygnus
Template XMM-Newton SAS Home Page
XMM-Newton Science Analysis System


asmooth (asmooth-2.10.2) [xmmsas_20030110_1802-5.4.1]

Method Adaptive smoothing Variance image Home Index

Meta Index / Home Page / Description / Adaptive smoothing


Template

The task has the facility both to write and to read a template image. The parameters writetemplateset, readtemplateset and templateset are relevant. The template image values represent the task's best estimate of the width of the gaussian required to achieve the desired SNR for each pixel.

NOTE that the definition of the template values is changed in versions 2.7 and later of asmooth. Versions 2.7 and above write a keyword FORMAT to the template files which enables the task to distinguish old- from new-format files. An old-format template file will be converted by the task to new format. This ought not to result in any differences in the smoothed image.

If readtemplateset is set `true' (the default is `false'), the task attempts to read its starting template from a file with a name specified by parameter templateset; if readtemplateset=`false', the task generates a new template internally. If writetemplateset=`true' (the default), this internal template is stored upon task completion.

The facility to store and reread a template image is valuable when it is desired to smooth several different input images in exactly the same way, for example during the creation of smoothed input images to colimplot. To see the benefit of this, consider the production of a hardness- or wavelength-coded colour image by colimplot. Several images in separate energy bands must be submitted to the task. A better-looking colour plot is produced if these are adaptively smoothed by asmooth before submission. But a source that is of high amplitude in one band may have a much lower amplitude in another. If just the input image values themselves were used to control the smoothing, a source in its low-amplitude band would be broader than in its high-amplitude band, resulting in chromatic aberration in the final colour image. Hence the desirability of employing the same user-supplied template to ensure an identical smoothing map for all three images. One might for instance do something like this:

asmooth inset=<all-band file> outset=<junk> readtemplateset=no
  writetemplateset=yes

asmooth inset=<band 1 file> outset=smoothed_1.ds readtemplateset=yes
  writetemplateset=no

asmooth inset=<band 2 file> outset=smoothed_2.ds readtemplateset=yes
  writetemplateset=no

etc.

colimplot insets='smoothed_1.ds smoothed_2.ds ..' etc.

Note that the values of readtemplateset and writetemplateset given in the first invocation of asmooth above are in fact the default values.


Method Adaptive smoothing Variance image Home Index

XMM-Newton SOC/SSC -- 2003-01-10