3. Image processing
The PHT Interactive Analysis (PIA) version 6.4 was used with the default settings to reduce the edited raw data created in November, 1996 via the off-line processing version 5.1 or 5.2. The top panel in Fig. 1 shows the results (hereafter called AAP map or signal) obtained from the PIA's Astronomical Analysis Processing (AAP). Checked patterns are evident in the C_90 AAP map. This is caused by drift in the responsivity of C100 detectors and unstable characteristics of the internal calibrators.
The top panel of Fig. 2 plots the AAP signal from C100 detector #5 as a function of time; each data point is a median mean of all the integration ramps taken at the respective raster point. This is one dimensional display of raster points. The AAP signal consists of high and low frequency fluctuations. The middle panel shows the smoothed signal obtained by filtering off the high frequency fluctuations from the AAP signal. For this filter, IDL's MEDIAN was first applied with WIDTH M. M is the number of raster points along the raster leg (i.e., 18 for C_90 and 27 for C_160 as given in Table 2). To remove small persisting ripples, SMOOTH was then applied with WIDTH . The bottom panel shows the drift corrected signal, which was obtained by dividing the AAP signal by the smoothed signal. This removes the low frequency fluctuations (detector drift) from the AAP signal. The bottom panel of Fig. 1, which was created from the corrected signal, demonstrates the drastic improvement for the C_90 map made by this drift correction processing. The AAP C_160 map is almost identical to that from the corrected signal. Comparison between AAP and corrected C_160 maps shows that the photometric error introduced from this correction is 4% or better and no extended brightness structures recognized within a area (see Fig. 1 in Kawara et al. 1997a) are affected. Fig. 3 shows the mosaiced maps of LHEX and LHNW, which are made from the corrected signal.
© European Southern Observatory (ESO) 1998
Online publication: July 7, 1998