CCD Superintendent

[MarkS]

If the green lines are the direction of elongation, there a some weird occurances of adjacent or overlapping stars with perpendicular elongation.
Could that be caused by imperfections with the mirrors or gas/dust in the area of the subject ?

The explanation is that the algorithm has been misled, probably because of noisy pixels or because a double star has been processed as a single entity.  Maybe it wasn't even a star in the first place.  In the end, the algorithm is taking a bunch of bright pixels, calling it a star and then calculating parameters for it.  Anything can go wrong and it usually does!

It's the general trend that is the important thing.

For instance, here are some statistics obtained by averaging out values for stars in different regions of the image - splitting the image into 5x5 areas:

Starcount
     110       96       94      103       66
      97      104       79       85       62
      99       65       76       60       59
      72       58       55       41       48
      70       52       54       45       44

FWHM (pixels)
    2.43     2.70     2.72     2.98     3.17
    2.58     2.68     2.82     3.06     3.34
    2.70     2.74     2.85     3.19     3.43
    2.78     2.92     3.10     3.38     3.57
    2.94     3.16     3.31     3.58     3.86

FWHM std dev (pixels)
    0.29     0.33     0.23     0.21     0.29
    0.29     0.27     0.19     0.25     0.24
    0.23     0.23     0.17     0.34     0.45
    0.36     0.36     0.26     0.20     0.48
    0.26     0.22     0.18     0.16     0.30

Elongation (pixels)
    0.52     0.56     0.54     0.69     1.01
    0.57     0.53     0.51     0.64     0.98
    0.50     0.40     0.45     0.66     0.88
    0.45     0.40     0.37     0.52     0.87
    0.46     0.28     0.30     0.49     0.99

Orientation of Major Axes (degrees)  (+ve is anti-clockwise from x-axis)
   -83.0    -78.1    -75.3    -73.8    -72.1
   -83.2    -75.2    -76.4    -81.6    -81.6
   -80.1    -80.4    -83.1    -89.5    -88.6
   -79.2    -77.1     88.0     83.2     80.1
   -71.3    -69.0     71.2     66.7     67.1

Mean Spread of Angle of Orientation (deg)
+/-13.4  +/-15.9  +/-10.5  +/- 8.6  +/- 6.6
+/-12.2  +/-14.8  +/-11.6  +/- 8.5  +/- 6.2
+/-13.1  +/-16.9  +/-13.2  +/-10.8  +/- 6.6
+/-15.1  +/-19.6  +/-19.4  +/-12.4  +/- 7.6
+/-19.9  +/-29.6  +/-21.8  +/-12.9  +/- 7.7

Major Axis FWHM (pixels)
    2.69     2.98     2.99     3.32     3.68
    2.86     2.94     3.08     3.38     3.82
    2.95     2.94     3.08     3.52     3.87
    3.01     3.12     3.29     3.64     4.01
    3.17     3.30     3.47     3.83     4.35

Minor Axis FWHM (pixels)
    2.17     2.42     2.45     2.64     2.67
    2.29     2.41     2.57     2.73     2.85
    2.46     2.54     2.63     2.86     2.99
    2.56     2.72     2.92     3.12     3.14
    2.71     3.01     3.16     3.34     3.36

[Mac]

Looking good,

[MarkS]

All the functionality I originally wanted is now in the prototype.  The big "windrose" in each panel indicates the overall direction of elongation of the stars in that panel.  The length of the main line in the windrose gives an indication of how significant the overall trend is - if the stars have fairly random directions then the line is very short.  The statistics within each pane relate to the averages of the stars within that pane.



As requested by Robert, I can now load FITS files (although I've only tested this on a Starlight Express sample).