SExtractor (Source Extractor) (full aka 'galactic' implimentation):
See
SExtractor home page with links to documentation for a full explanation of how sources are recognised, separated, measured etc.
SExtractor was written for optical images but it can be used on any suitable FITS pixel images containing compact sources (typically 3-10 pixels in size) which can be approximated by ellipses. Non-standard parameters may be needed for e.g. background estimation and flux density scaling for non-optical images.
The details of source extraction are controlled by a configuration file e.g.
myfile.sex. Some of the values may be overwritten by other inputs, see below. This allows a single configuration file to be used for many images taken by the same instrument but with different sensitivities etc. The output measurements are specified in a parameter file e.g.
myfile.param and are in
VOTable format.
The following options are available (see SExtractor documentation for more explanations):
| Analysis threshold1 | Multiple of sigma_rms for source detection threshold |
| Configuration file2 | Link to .sex file e.g. myfile.sex |
| Detection image2,3 | Link to a FITS image. The detection image is used to locate sources according to the criteria in the Configuration file. |
| detection filter2,5 | The spatial filter (.conv) used in detection e.g. default.conv |
| Input image band | Band label for measurement image |
| Magnitude zero point1,4 | For images in detector units, this should be the effective zero point for conversion to standard magnitudes or physical units. This requires that the correct information is in the FITS header. For images in magnitudes this should be the standard zeropoint for conversion to physical units, if required. |
| Parameter file2 | Link to .param file myfile.param |
| Measurement image 2,3 | User supplies a link either to a MySpace file or a URL. The output source parameters are measured from this image, at the positions found from the detection image. |
| Seeing1,4 | Required for star-galaxy separation (or equivalent, e.g. beam size for radio images, for separating point and slightly extended sources). |
| Derived output catalogue | User supplies MySpace file name for output measurements. |
Notes:
- All the step box entries can contain variables controlled by a script
- These parameters can be specified in the .sex file, in which case use the rightmost check box on the line for the parameter to delete the line from the step box. Otherwise the entry in the step box will be used.
- This should be a reference either to an http address e.g. INT-WFS examples (see intwfs.sex, default.conv, intwfs_small.param) or to a file which you have uploaded into your MySpace
- The Detection and Measurement images can be the same, making it identical to using SExtractor in single image mode;
- The Detection and Measurement images must be identical in pixel and world coordinates (e.g. both 1024x2048, reference pixel at (512, 1025) corresponding to the same RA and Dec).
- This information may be in the FITS header or observing log or web page. Example: ZeroPoint and Seeing for INT-WFS images
- If you do not want to use a spatial filter, use the rightmost check box on the line for the parameter to delete the line from the step box and set FILTER N in the .sex file but do not leave FILTER_NAME blank, leave as default.conv.
Tool wrapped by
MartinHill and
PaulHarrison
--
AnitaRichards - 14 Mar 2005