Next: What Data Frames can
Up: Object Search and Classification
Previous: Object Search and Classification
INVENTORY has been originally designed as a medium speed and medium
accuracy universal program for finding, classifying, and investigating
astronomical objects on two-dimensional image frames in a way that is
as automatic as possible. In a course of further development, both speed
of execution and attainable precision have been improved, and the
package differs from other related programs mainly by its tendency
to minimise the amount of time spent by a user on interactive work at
the terminal. Though it can be used for most of the relevant applications,
INVENTORY is best suited for analysing numerous similar frames, like
in cases of surveys or variable stars observations. It is also
convenient for a first look preview of the material that will be
analysed with more elaborate methods. It is not recommended for users
who are not willing to give up the possibility of interactive control.
A major requirement imposed on INVENTORY was that it should be
able to classify detected objects into stars, galaxies, and
image defects.
Note
While it is possible to run INVENTORY for the first time
without many preparations, achieving good results requires some
experience. The program runs in an automatic mode when some
parameters are set to proper values. Different applications may
require different values for these parameters.
Using INVENTORY with a new kind of material requires some
preparation and trial runs are
necessary for adjusting the parameters. The time spent on tuning up the
program is well paid off when there is a sufficiently large number
of objects and/or frames so that the use of other techniques
would be much more time consuming. The tuning up may be difficult in a
crowded field with many overlapping objects, or in a case where there are
big bright galaxies. On the other hand the use of INVENTORY should
be relatively easy when the investigated field is populated mainly by not
too densely packed stars.
These photometric packages which aspire to high accuracy of results
have to solve the problem of deblending of overlaping images. In its
base mode INVENTORY does it in a fast but approximate way. A
somewhat more
accurate deblending of stellar objects using one-dimensional point spread
function is now optional. Two dimensional point spread function fit,
its extension to strongly undersampled images, and more precise
deblending of galaxies are in preparation.
The program is functionally divided into three discrete steps. It will be
possible to perform these steps using the commands: SEARCH/INV, ANALYSE/INV,
and CLASSIFY/INV. The additional commands
SHOW/INV and SET/INV serve for
displaying and updating values of the INVENTORY keywords.
Before using INVENTORY commands it is necessary to give command
SET/CONTEXT INVENT. It may be convenient
to include this command into the login.prg file.
-
- The (optional) SEARCH/INV command prepares a preliminary list
of objects.
The SEARCH/INV can be omitted once we have a list of
objects with accurate positions in a MIDAS table format.
-
- The ANALYSE/INV command evaluates and updates an input list of
objects. It can be used in a VERIFY or NOVERIFY mode.
In VERIFY mode, the ANALYSE/INV command verifies the
used table of objects. Some entries are deleted but usually a
larger number of new ones are added. The objects positions are improved.
In NOVERIFY mode this verification process is omitted.
In both modes the ANALYSE/INV command calculates several image
parameters, which can be used as final results and/or as input
to the CLASSIFY/INV command.
-
- The CLASSIFY/INV command uses the output table produced by the
ANALYSE/INV command for dividing the objects into stars,
galaxies and
spurious objects. It accepts only input of MIDAS table files that
have been produced by ANALYSE/INV.
Next: What Data Frames can
Up: Object Search and Classification
Previous: Object Search and Classification
Petra Nass
1999-06-15