Detection ignores images with too many stars

[markmac99]

There's an upper threshold (normally 400) above which RMS rejects an image for having too many stars.

However, there's no notification to the user, because the code uses print() to emit a warning, and the output of print() is only visible onscreen. Once the Pi reboots, the message is lost.

The code should log() the info to the RMS logfile, and at the end of a run, if more than a (configurable) amount of frames were ignored, then some further warning should be emitted as data is being lost.

For example, perhaps the astrometry report image could be overwritten with "TOO MANY STARS" in place of "NO MATCHED STARS", or the "observing_periods" report could be marked to show that frames were ignored (red shading?).

Note: this problem can be caused by noise on the image, which might be due to the camera AGC setting being too high. We could therefore include a suggestion to examine and adjust the AGC. Cameras are normally set with AGC=60, which has been tested and found to be reasonable by Istrastream and UKMON, but in a dark site or a noisy camera then 20-30 might be better as suggested by @satmonkey. We should certainly make sure there's flexibility.

[markmac99]

Updated ExtractStars.py to log to the logfile. The message "Too many candidate stars to process! x/y" will now appear in the logs. Adding a more prominent warning is harder to do and will require further workl.

[markmac99]

Added logging to ExtractStars.py, using the standard RMS logger. Sample output 2023/07/30 03:17:32-INFO-ExtractStars-line:186 - extracted 68 stars from FF_UK0006_20230730_031713_995_0593408.fits 2023/07/30 03:17:41-INFO-ExtractStars-line:186 - extracted 63 stars from FF_UK0006_20230730_031724_249_0593664.fits 2023/07/30 03:17:50-INFO-ExtractStars-line:186 - extracted 61 stars from FF_UK0006_20230730_031734_483_0593920.fits