3 The spectrum detection procedure DETSP

The above-described algorithm is implemented in a procedure named DETSP, part of a general context OBJPR (Pasian et al. 1997) for objective prism image processing, developed in collaboration by the Astronomical Observatory of Trieste and the Astronomical Institute of the National Observatory of Athens. The context is worked out under the Munich Image Data Analysis System (MIDAS 96NOV 1996).

The DETSP takes as input an image frame from a digitized spectral plate and particular parameters for the spectral images (dispersion, length, width). The subframe overlapping can be applied to both axes (defaulted), only down the wavelength axis, or skipped altogether. Optionally, working image displaying and/or online graphics can be switched on. Output is a MIDAS format table with the detected spectral images positions. It is suitable for use by the procedures of the OBJPR, say for extraction the detected spectral images. The processing is carried out in four sequential stages:

${\it a.~Image~frame~preprocessing}$
The whole image frame is filtered by a sequence of median and smoothing filters. A grid of subframes (see Sect. 2.2) is fixed on the filtered image, according to the overlapping mode.

${\it b.~Subframe~signal~processing}$
Each one of the fixed subframes is processed applying the signal detection algorithm (see Sect. 2.1). The detected spectral positions are saved in table.

${\it c.~Detection~table~processing}$
There are possible double detections of spectra near the edges of neighboring subframes. For this reason, the table of detected spectra is now processed to remove the doubling. It is sorted as well.

${\it d.~Detection~fine~adjustment}$
The signal processing approach is used again. Now, as many subframes are fixed as the number of detected spectra. The subframes are narrower and each one includes a particular detected spectrum image. This leads to fine adjustment of the position. The adjusted position table is finally sorted.